Article Text
Abstract
Objective Since approval in Crohn’s disease (CD) of risankizumab, there has been widespread use. Real-world data are, however, limited and our aim is to address that gap.
Design/method We performed a retrospective, observational study of risankizumab use in patients with CD starting treatment between January 2021 and January 2023 at two UK centres. Clinical activity, biochemical and faecal biomarkers were measured at baseline, weeks 4, 12, 28 and 52. The primary outcome was clinical response at weeks 12, 28 and 52.
Results 53 patients (51% women); median (range) age 40 years (20–70); median disease duration 15 years (6–52). Clinical response was observed in 33% (n=14/42), 45% (n=17/38) and 52% (n=13/25), and clinical remission in 31% (n=13/42), 40% (n=15/38) and 44% (n=11/25) at weeks 12, 28 and 52, respectively. Median C reactive protein decreased from 12 mg/L (IQR: 4–30; n=50) at baseline to 6 mg/L (IQR: 2–16; p=0.03 vs baseline; n=49) at week 12, 3 mg/L (IQR: 2–8, p=0.003; n=44) at week 28 and 3 mg/L (IQR 1–4, p=0.007; n=31) at week 52. Median faecal calprotectin concentration was 668 µg/g (IQR: 246–1098; n=32) at baseline, 298 µg/g (IQR: 176–546, p=NS; n=21) at week 12, 358 µg/g (IQR: 133–622, p=0.03; n=14) at week 28 and 63 µg/g (IQR: 38–120, p=0.007; n=12) at week 52.
12 out of 18 patients discontinued corticosteroids at week 12, 16 by week 28 and 18 by week 52. Four major adverse events—three elective and one emergency surgery—were recorded.
Conclusion Risankizumab is effective in a refractory real-world population with CD.
- INFLAMMATORY BOWEL DISEASE
- IBD
- IBD CLINICAL
- CROHN'S DISEASE
Data availability statement
Data are available on reasonable request.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
Risankizumab is a monoclonal antibody inhibiting interleukin-23p19. Two induction and one maintenance trial (ADVANCE, MOTIVATE and FORTIFY) demonstrated efficacy in moderate-to-severe Crohn’s disease (CD).
WHAT THIS STUDY ADDS
This work presents one of the longest follow-up studies of a real-world cohort with CD refractory to vedolizumab, ustekinumab and at least one tumour necrosis factor-alpha antagonist, treated with risankizumab. Our data suggest a clinically relevant and significant impact on clinical disease activity and biomarkers, after induction and out to week 52 along with a marked steroid-sparing effect.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
This study represents an objective and generalisable assessment of risankizumab’s effectiveness in the real-world setting of patients with CD with previous multiple biologic failures.
In practice, clinicians should consider risankizumab in patients with moderate-to-severely active CD despite use of prior biologic therapies as well as in those who are steroid dependent.
Introduction
Crohn’s disease (CD) is a chronic inflammatory disease in which anti-tumour necrosis factor (anti-TNFα) agents were the mainstay of treatment for many years.1 2 However, a significant number still do not respond to these therapies and, over the preceding half a decade, the understanding of new pathways involved in the pathogenesis of inflammatory bowel disease (IBD) has led to development of new drugs targeting different pathways in the inflammatory cascade.3
Risankizumab is a new generation humanised monoclonal antibody that selectively binds to, and inhibits, interleukin-23p19.4 It has previously been approved in inflammatory diseases other than IBD, including psoriatic arthritis and plaque psoriasis.5 Data from the phase III clinical trials (ADVANCE, MOTIVATE and FORTIFY) demonstrated efficacy in induction and maintenance of remission in moderate-to-severe CD,6–8 resulting in Food and Drug Administration (FDA) approval in June 2022 and European Medicines Agency (EMA) approval in November 2022.9
Prelicence access to risankizumab was granted through a preapproval access programme (PAAP), to patients with moderately-to-severely active CD who had an inadequate response to, lost response to, were intolerant of or who had contraindications to infliximab, adalimumab, vedolizumab and ustekinumab. Subsequently, from April 2022, an early access to medicines scheme (EAMS) permitted access to a similar cohort, the only difference being that this scheme granted supply to patients who experienced inadequate/loss of response or intolerance to vedolizumab, ustekinumab and a single anti-TNFα agent. Given the paucity of published data on the use of risankizumab for this indication in a non-trial environment, our aim was to investigate the effectiveness of risankizumab for treatment-refractory CD in a real-world setting.
Methods
Inclusion criteria and study population
We performed a retrospective cohort study by reviewing prospectively maintained clinical records for all patients with CD starting risankizumab between January 2021 and January 2023, at two tertiary UK IBD centres. Eligibility for risankizumab consisted displaying symptoms and/or objective signs of active inflammation (elevated C reactive protein (CRP) and/or faecal calprotectin (FC) and/or radiologic and/or endoscopic evidence of disease activity), as well as previous history of biologic exposure as described above.
Induction with three intravenous doses of 600 mg of risankizumab was given at weeks 0, 4 and 8 followed by 8-weekly subcutaneous, maintenance dosing from week 12. Eight subjects received the initial PAAP maintenance dose of 180 mg, with six of the subjects who started at 180 mg being increased to 360 mg based on clinical and objective assessments of disease activity. The remaining 45 received 360 mg maintenance dosing throughout.
Data collection
Patient demographics, CD history (disease duration, age of onset and surgical history) and disease distribution and phenotype (Montreal classification) were recorded alongside previous and current treatment history. Clinical disease activity was assessed using Harvey-Bradshaw Index (HBI) and IBD-Control Score was recorded at one site; biochemical disease activity assessments were made using CRP and FC. Data were collected at baseline, weeks 4 (except in the case of FC), 12, 28 and 52 at the time of medication administration. Concomitant medication use was recorded at each visit, as were adverse effects. In addition, endoscopic or imaging investigations performed at/prior to baseline and during the follow-up period were recorded.
Key outcome measures
Our primary outcome was clinical response (defined as a reduction of three or more points in HBI compared with baseline, or achievement and persistence of HBI<5) at weeks 12, 28 and 52. We also evaluated clinical remission (defined as HBI<5).10–12 Patients with stomas (n=10) were excluded from HBI analyses.
Secondary outcomes included corticosteroid-free clinical remission (CSFR) defined as HBI<5 in the absence of concurrent corticosteroid administration. Additional secondary outcomes were biochemical remission by blood biomarkers (CRP≤5 mg/L), and biochemical remission according to stool biomarkers (FC≤200 µg/g), at weeks 12, 28 and 52, as well as combined biochemical remission, defined by CRP≤5 mg/L and FC≤200 µg/g.
Patients with HBI<5 at baseline (potentially not in keeping with active disease based on clinical criteria) were automatically designated not to have achieved clinical response or (steroid free) clinical remission, regardless of eventual absolute (or change in) HBI at follow-up. Individuals with CRP<6 mg/L at baseline (again, likely in keeping with biochemically non-active disease) were automatically classified as not having achieved biochemical remission, regardless of the subsequent value.
Statistical analysis
Data analysis was carried out using GraphPad Prism V.9.
Results
Patient characteristics
53 patients were enrolled between 2 centres (table 1). Median age and duration of CD were 40 (20–70) years and 15 (6–52) years, respectively. Over half of participants had stricturing phenotype (51%; n=27/53), and nearly one-fifth had penetrating disease (17%; n=9/53). The majority had ileocolonic CD (79%; n=42/53) and nearly half had perianal involvement (47%; n=25/53). Prior resection was noted in 62% (n=33/53) of patients, and 10 individuals had stomas in situ. All 53 participants had exposure to three classes of prior biologic therapy, with 94% (n=50/53) having previously been treated with both infliximab and adalimumab, the remainder having received one of the two.
Clinical response and remission
At baseline, 65% (n=28/42) of patients had clinically active disease (as defined by HBI≥5).
Assessing our primary outcome—clinical response was observed in 33% (n=14/42) at week 12, 45% (n=17/38) at week 28 and 52% (n=13/25) at week 52, and clinical remission was achieved by 31% (n=13/42) at week 12, 40% (n=15/38) at week 28 and 44% (n=11/25) at week 52 (figure 1).
Of those patients who had not achieved clinical response or remission by HBI criteria at week 12, seven individuals subsequently obtained clinical response—five at week 28 (with no proceeding worsening of HBI in all three cases where data was available) and two at week 52; all of these seven apparent late responders except also met HBI criteria for clinical remission—four at week 28 (sustained for the two where data was available) and two at week 52.
Baseline HBI 8 (IQR: 4–11 for n=42) fell to 4 (IQR: 1–4, p=0.0011 vs baseline for n=31) at week 4 after a single induction dose; HBI was 3 (IQR: 1–6, p=0.0006 vs baseline for n=36) at week 12, at the end of induction. During maintenance therapy, at week 28 median HBI was 3 (IQR: 1–6, p=0.0004 vs baseline for n=33) and at week 52 it fell further to 2 (IQR: 1–4, p<0.0001 vs baseline for n=20) (figure 2A, table 2).
Among the 14 patients with HBI<5 at baseline, HBI remained<5 throughout the study period aside from in three individuals. Among these three, one had an HBI of 6 at week 12 on risankizumab that fell to 3 and then 1 by weeks 28 and 52, respectively; another had HBIs of 6 and 10 at week 12 and 28, but this also dropped to 1 by week 52; the third had an HBI of 12 at week 28 and risankizumab was discontinued in this patient prior to the week 52 dose due to perceived lack of efficacy.
Biochemical measures of response
At baseline, 62% (n=33/53) had CRP>5 mg/L and 45% (n=24/53) had FC>200 µg/g. Biochemical remission by CRP (≤5 mg/L) was achieved in 24% (n=8/33) at week 12, 36% (n=12/33) at week 28 and 36% (n=12/33) at week 52. Biochemical remission by FC (≤200 µg/g) was observed in 25% (n=6/24) at week 12, 17% (n=4/24) at week 28 and 38% (n=9/24) at week 52, while combined biochemical remission was seen in no individuals at week 12, but 16% (n=6/38) at both week 28 and 52 (figure 1).
Median CRP decreased from 12 mg/L (IQR: 4–30 for n=50) at baseline to 7 mg/L (IQR: 2–15, p=0.003 vs baseline for n=49) at week 4, 6 mg/L (IQR: 2–16, p=0.029 vs baseline for n=49) at week 12, 3 mg/L (IQR: 2–8, p=0.003 vs baseline for n=44) at week 28 and 3 mg/L (IQR: 1–6, p=0.007 vs baseline for n=31) at week 52. CRP values were missing for four individuals at each of weeks 4 and 12, one at week 28 and one at week 52 (figure 2C, table 2).
Median FC decreased from 668 µg/g (IQR: 246–1098 for n=32) at baseline to 298 µg/g (IQR: 176–546, p=0.597 vs baseline for n=21) at week 12, 358 µg/g (IQR: 133–622, p=0.026 vs baseline for n=14) at week 28, and to 63 µg/g (IQR: 38–120, p=0.007 vs baseline for n=12) at week 52 (figure 2D, table 2).
Steroid-free disease control
Corticosteroids were prescribed at baseline in 34% (n=18/53) of patients. Of these, 67% (n=12/18) had discontinued within the first 12 weeks of treatment with risankizumab and a further 22% (n=4/18) discontinued during maintenance treatment. 11% (n=2/18) remained on treatment at week 28 one of whom discontinued steroids by week 52. One patient initiated corticosteroids between weeks 12 and 28 follow-up visits but discontinued steroid therapy by week 52.
CSFR (defined by HBI<5 in absence of concurrent corticosteroid administration) was achieved in 29% (n=12/42) at week 12, 37% (n=14/38) at week 28 and 44% (n=11/25) at week 52 (figure 1).
Treatment persistence
At the end of a follow-up of 515 days (IQR: 332–653), risankizumab persistence was 74% (n=39/53). While all patients aside from one remained on risankizumab at week 12, five individuals discontinued therapy during maintenance prior to the 28-week follow-up visit and a further eight prior to the 52-week follow-up (figure 3).
Reasons for drug discontinuation were primary non-response (23% (n=12/53)), nine of whom switched to alternative therapy (seven to upadacitinib, one to golimumab and one to a trial using autologous regulatory T cells) and three of whom had elective surgery. One further patient underwent resectional surgery during follow-up on an emergency basis (representing the only case of CD-related hospitalisation throughout the study period). One patient discontinued risankizumab after relocating outside of the UK.
Patient-perceived disease control
Patient-perceived disease control (IBD Control) significantly increased from 4 (IQR: 2–8; n=30) at baseline to 11 (IQR: 8–14, p<0.0001 vs baseline; n=30) at week 4 and 10 (IQR: 6–14, p=0.0002 vs baseline; n=29) at week 12. During maintenance treatment, the week 28 median score was 13 (IQR: 10–14, p=0.0069 vs baseline; n=15) and week 52 was 14 (IQR: 10–15, p=0.3774 vs baseline; n=3) (figure 2B, table 2).
Safety
Nine adverse events (AEs) in eight people were recorded—with respect to major events, one patient required emergency surgery and three required elective surgery during treatment, before which risankizumab was discontinued; five mild events were recorded (headache, flu-like symptoms, chest pain, bloating, nausea and palpitations), none necessitating drug discontinuation.
Discussion
Our study is one of only a handful of publications presenting real-world evidence to demonstrate the benefit of risankizumab, not only as an induction agent but also during maintenance up to 1-year follow-up. We present a sizeable cohort demonstrating risankizumab persistence of 74% at a year in a highly treatment-refractory population. A third and a half of patients demonstrated clinical response post induction and during maintenance, respectively; and, caveating conclusions in view of missing data, a robust biochemical response was evidenced by CRP of≤5 mg/L and FC≤200 µg/g in almost 40% of participants at the end of maintenance.
Further, these findings were observed in the context of no new safety signals. In a comparable population by way of phenotype, a multicentre French study carried out by Groupe d’Etude Thérapeutique des Affections Inflammatoires Digestives (GETAID)8 demonstrated a similar rate of clinical remission (using the same definition)—39%—and similar improvement in CRP from baseline—8.1 (3.0–30.4) mg/L to 5.0 (2.1–15.0) mg/L at week 12 (p=0.019) vs 12 mg/L (IQR: 4–30) at baseline to 6 mg/L (IQR: 2–16, p=0.0290) at week 28. Our dataset, while smaller, demonstrates ongoing effectiveness up to 1 year after starting treatment. This is mirrored by a recent update published in abstract form by the same group that reported an almost identical rate of clinical remission to our own (based on the same HBI-based criteria) of 50% (65/131) at 1 year.13
Among the two phase III induction trials, ADVANCE and MOTIVATE, 45% and 42% of patients achieved clinical remission at week 12 with 4 weekly intravenous administrations of 600 mg of risankizumab.12 While, in the placebo-controlled maintenance trial FORTIFY (participants were randomised to 8 weekly 180 mg or 360 mg subcutaneous maintenance or placebo) coprimary endpoints of clinical remission (CD activity index<150 in the USA, and patient outcomes of stool frequency and abdominal pain scores in Europe) and endoscopic response were achieved by 51.8% (USA)/52.5% (Europe) and 46.8%, respectively, at 52 weeks.8 Additionally, risankizumab returned similar numbers of overall AEs to placebo, with low rates of severe and serious AEs in these trials.5 14 The findings of our study therefore reflect the randomised controlled trial data, suggesting that risankizumab is safe as well as effective in both inducing and maintaining clinical remission in a population of patients with moderate-to-severely active CD with previous biologic failure.
In our study, the eligibility criteria for PAAP and EAMS dictated that our patient cohort was a highly treatment-refractory population, with 100% having been exposed to ustekinumab and vedolizumab and at least one of adalimumab or infliximab. By comparison, 58% of patients in ADVANCE had experienced previous biologic intolerance/inefficacy and, while all patients in MOTIVATE were biologic experienced, only 53% had been exposed to more than one agent, and just 19% had received ustekinumab. Additionally, our participants displayed particularly aggressive disease characteristics; over half had stricturing disease and 62% had undergone previous luminal surgery (62%). Furthermore, 47% had perianal involvement and nearly one-fifth had penetrating disease. In spite of this, we have shown clinically meaningful and significant decrease in median HBI by weeks 12 and 28 with benefit for many patients continuing to week 52. Corresponding improvements in biomarkers were also demonstrated.
As with the GETAID study,8 13 we also demonstrated marked steroid-sparing benefit in risankizumab-treated patients. Although over a third of our participants were receiving steroids at baseline, the majority had already discontinued by the end of induction, and the rest by the end of 52 weeks with a resulting CSFR of over 40%.
As previously noted, in ADVANCE and MOTIVATE the AE rates were comparable between placebo and risankizumab groups. However, data from the open-label extension of the Phase II trial15 (n=65), in which a lower maintenance dose of 180 mg 8 weekly was used, reported minor infections (nasopharyngitis, gastroenteritis, lower respiratory tract infection, oral herpes and urinary tract infection) as the most common AEs (73.8%), alongside hypersensitivity reactions (24.6%) and hepatic events (9.2%); serious AEs (though no deaths) were observed in 35% of participants (including viral gastroenteritis, peritonitis, intestinal stenosis, and postprocedural complications), resulting in six participants discontinuing risankizumab. Acknowledging that we cannot directly compare AEs between clinical trials and a real-world experience study such as our own, we demonstrated no malignancies or major cardiovascular events, in line with existing safety data. In support of these findings, one of the few other real-world cohort studies on risankizumab (collecting data from six centres in Belgian) reported no safety issues among their 69 patients. While four study participants in our cohort did require intestinal surgery, the majority of AEs recorded were minor and non-treatment modifying. This suggests that risankizumab is, on the whole, safe and well tolerated as a treatment in the short-to-medium term.
We acknowledge several important limitations of our study. First is its retrospective design, which has the potential to result in a range of possible biases. Second is the subjective nature of the clinical disease activity scores used to derive our primary and secondary outcomes—this was a conscious decision given that they are typically already measured as part of routine clinical care and that, for that reason precisely, they remain relevant to clinicians and patients. We attempted to address this subjectivity with the complimentary objective measurements of FC (interpretability of which was unfortunately hindered by missing data) and CRP; however, we do appreciate the increasing realisation that achieving endoscopic targets is an additional important aspect of adequate disease control that this study does not address robustly. Finally, this study lacked an especially systematic approach to collecting AE data, which limits the conclusions regarding these.
Conversely, this study has some positives. It was a dual-centre study of consecutively treated patients with (in most cases) clinically active CD at the time of drug initiation. And, for the majority of patients, the licensed dosing regimen was used, increasing the generalisability of the findings. Further, given our fairly rigorous approach to ensuring active disease at baseline as a prerequisite to classifying cases as having subsequently achieved remission—where HBI<5 or CRP<6 mg at baseline, cases were automatically designated as not having achieved clinical response/remission or biochemical remission, respectively—the study design ensures that the treatment effect of risankizumab is not overestimated.
In conclusion, we present a cohort of patients with highly refractory CD in whom meaningful benefit of treatment with risankizumab was demonstrated both after induction and during maintenance in a real-world setting. Furthermore, no new safety signals were identified.
Summary
This retrospective real-world cohort analysis of patients with CD refractory to multiple biologics suggests a positive impact of risankizumab on clinical disease activity, patient perceived disease control and biochemical disease activity, measured after 12 weeks of induction therapy, as well as at weeks 28 and 52 during maintenance.
Data availability statement
Data are available on reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
Not applicable.
Acknowledgments
The completion of this study and production of this paper was greatly assisted by the efforts of
Mark Sephton (Medical Science Liaison Manager, AbbVie UK).
Footnotes
BZ and BG are joint first authors.
CWL and MAS are joint senior authors.
X @zare_benjamin, @SamaanMark
Contributors Guarantor of the article: MAS. Development of study concept and design: BZ, BG, ES, PD, CWL and MAS. Study supervision: CWL and MAS. Acquisition, analysis and interpretation of the data: BZ, BG, NL, ES, PD and SL. Statistical analysis: BZ and RJD. Drafting of the manuscript: BZ, BG, CWL and MAS. Critical revision of the manuscript for important intellectual content: RJD, SL, SR, SHCA, JM, PMI, CWL and MAS.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests BZ—education bursary: Dr Falk. BG—advisory fees: Galapagos and Abbvie; lecture fees: Abbvie, Jansen, Takeda, Pfizer and Galapagos. CWL—advisory fees: Abbvie, Janssen, Takeda, Pfizer, Galapagos, Bristol Myers Squibb, B.I., Sandoz, Novartis, GSK, Gilead, Vifor Pharma, Dr Falk, Trellus Health and Iterative Scopes; lecture fees: Pfizer, Janssen, Abbvie, Galapagos, MSD, Takeda, Shire, Ferring, Hospira and Dr Falk. MAS—Advisory fees: Janssen, Takeda, Sandoz, Samsung Bioepis, Galapagos, AbbVie, Bristol-Myers Squibb, Pfizer, Tillotts; lecture fees: Bristol-Myers Squibb, Janssen, Takeda, MSD, Falk, AbbVie, Galapagos.
Provenance and peer review Not commissioned; externally peer reviewed.
Linked Articles
- Highlights from this issue