Single-center Experience With Upadacitinib for Adolescents With Refractory Inflammatory Bowel Disease

Abstract

Background

Upadacitinib (UPA) is a novel selective JAK inhibitor approved for adults with ulcerative colitis (UC) and with positive phase 3 data for Crohn’s disease (CD). Pediatric off-label use is common due to delays in pediatric approvals; real-world data on UPA are needed to understand the safety and effectiveness in pediatric IBD.

Methods

This is a single-center retrospective case series study of adolescents (12-17 years) with inflammatory bowel disease IBD on UPA. The primary outcome was postinduction steroid-free clinical remission (SF-CR) defined as Pediatric UC Activity Index (PUCAI) or Pediatric CD Activity Index (PCDAI) ≤10. Secondary outcomes include postinduction clinical response (decrease ≥12.5 in PUCAI/PCDAI), postinduction C-reactive protein (CRP) normalization, 6-month SF-CR, and intestinal ultrasound response and remission. Adverse events were recorded through last follow-up.

Results

Twenty patients (9 CD, 10 UC, 1 IBD-U; 55% female; median age 15 years, 90% ≥2 biologics) were treated with UPA for ≥12 weeks (median 51 [43-63] weeks). Upadacitinib was used as monotherapy in 55% and as combination with ustekinumab and vedolizumab in 35% and 10%, respectively. Week 12 SF-CR was achieved in 75% (15/20) and 80% (16/20) with CRP normalization. About 3/4 (14/19) achieved SF-CR at 6 months. Adverse event occurred in 2 patients (10%): Cytomegalovirus colitis requiring hospitalization and hyperlipidemia requiring no treatment. In the 75% with ultrasound monitoring, response and remission were achieved in 77% and 60%, respectively.

Conclusion

While awaiting pediatric registration trials, our data suggest that UPA is effective in inducing and maintaining SF-CR in adolescents with highly-refractory IBD with an acceptable safety profile.

Graphical abstract

Graphical Abstract.

What is already known?

Upadacitinib, a selective JAK inhibitor, is effective in inducing and maintaining remission in adults with ulcerative colitis and Crohn’s disease.

What is new here?

This case series presents novel data on the effectiveness and safety of upadacitinib in adolescent patients with IBD.

How can this study help patient care?

Pediatric providers are often required to prescribe therapies off-label before pediatric registration trials complete. These data will inform pediatric providers and aid in the insurance approval process.

Introduction

Biologic and small molecule therapies are the mainstays of medical management of inflammatory bowel disease (IBD).¹ Unfortunately, while the therapeutic armamentarium is growing for adult IBD and thereby expanding options for successive therapeutic trials, long delays between drug approvals for adult and pediatric populations have led to a static number of treatment options for children. Infliximab and adalimumab, both tumor necrosis factor (TNF) inhibitors, remain the only approved advanced IBD therapies for children under the age of 18. This results in the commonplace practice of off-label prescribing,² and pediatric physicians rely on initial observational data to inform their management and aid in insurance approvals. The 2 most common concerns of pediatric providers when considering off-label therapies revolve around safety and dose; importantly, concerns of both over- and underdosing.

Upadacitinib (UPA) is a novel selective Janus kinase inhibitor (JAKi) that represents an important newcomer to the IBD armamentarium.³ It was recently approved by the FDA for patients 18 and older with ulcerative colitis (UC)⁴ and Crohn’s disease (CD).⁵ Most notably, it has preserved effectiveness in the face of multiple prior advanced therapy failures, thus offering a solution to an important gap in clinical care.^4,6,7 Further, rapid clearance of biologic therapies is particularly common in the pediatric population,^8,9 and given the nature of small molecule therapies, patients who have experienced limited exposure and effect to advanced therapies due to rapid biologic clearance may benefit from a change to a JAKi.¹⁰

In adults with both UC and CD, the oral induction dose is 45 mg daily, with reduction in maintenance to 30 mg or 15 mg daily. In pediatrics, UPA is FDA-approved for use in atopic dermatitis (AD) down to the age of 12, dosed either at 15 mg or 30 mg daily, with a favorable safety profile.¹¹ The adult IBD induction dose of 45 mg daily, thus, diverges from existing data on UPA in teenagers with AD, and adverse events (AEs) have been reported to be dose-dependent with UPA,^4,5 making the safety of this higher dose of interest.

There is currently a paucity of real-world data on the effectiveness and safety of UPA in pediatric IBD, with only a single promising case report on its use in a 12-year-old patient with CD.⁶ As with other adult-approved IBD therapies, published pediatric real-word effectiveness data are critical to help guide the management of patients with refractory pediatric IBD and to provide the data needed for the health plan authorization process. Thus, we aim to present herein our initial experience with UPA in adolescent IBD.

Methods

Study Population

This is a retrospective case series study of adolescents with IBD ages 12 to 17 prescribed UPA at a single tertiary US IBD center between January 1, 2020, and May 3, 2023. Patients were identified by querying a health-system-wide data warehouse, which contains electronic health record data for over 11 million patients, for patients younger than 18 years old with International Classification of Diseases (ICD) codes for CD, UC, and/or IBD-unclassified (IBD-U), and an electronic prescription for UPA.

Data Collection

All data were obtained through retrospective review of the electronic medical records. Data were gathered on demographics, disease subtype, Paris classification,¹² indication for UPA, time on UPA (days), number of biologic failures, prior tofacitinib use, concomitant biologic use, corticosteroid use, and presence or absence of active joint involvement (eg, sacroiliitis, arthritis, arthralgias). Clinical scores (Pediatric UC Activity Index [PUCAI] or Pediatric CD Activity Index [PCDAI]) and C-reactive protein (CRP) were gathered. Adverse events (AEs), including serious AE (SAEs), were collected on all patients, specifically assessing for infection, hospitalization, cardiovascular events, dyslipidemia, leukopenia, anemia, abnormal liver function tests, or malignancy.

Intestinal Ultrasound

Intestinal ultrasound (IUS) examinations were performed with a Samsung RS85 Prestige (Samsung, South Korea) located in the IBD center utilizing a low frequency convex probe (3-10MHz) for global assessment and a high frequency, linear probe (2-14MHz) for more detail as appropriate. Intestinal ultrasound examinations were performed by one of 3 certified international bowel ultrasound (IBUS)-trained gastroenterologists. No bowel preparation or fasting was required prior to examination. A standardized approach was undertaken for each IUS examination, recording these IUS parameters: Bowel wall thickness (BWT), bowel wall hyperemia (BWH) measured by color doppler signal, with a velocity rate of ±5.2 cm/s, and graded according to the modified Limberg score,¹³ mesenteric inflammatory fat, and lymphadenopathy.¹⁴ Intestinal ultrasound scores were calculated as appropriate at baseline and postinduction. Intestinal ultrasound defined the most affected bowel segment at baseline as the segment with the largest BWT. Ultrasound response was defined as ≥25% decrease in BWT, 2 mm absolute decrease in BWT, or 1 mm decrease in BWT accompanied by a 1 point decrease in modified Limberg score, all in the most affected segment. Ultrasound remission was defined as BWT <3 mm and absent BWH.¹⁵

Outcome Measures and Analysis

Descriptive statistics summarized the data. The primary outcome of this study was postinduction steroid-free clinical remission (SF-CR), which was defined as PUCAI ≤10 following 8 weeks of induction or PCDAI ≤10 following 12 weeks of induction, all in the absence of steroids. Secondary outcomes included postinduction steroid-free clinical response (defined as a decrease from baseline of ≥12.5 points in PCDAI/PUCAI in absence of steroids), postinduction CRP remission (defined as normal CRP based on lab parameters), postinduction resolution of joint pain in those with joint pain at baseline, IUS response, and remission in a subset of patients who had IUS at baseline and postinduction, and SF-CR at 6 months. All AEs and SAEs were described through last follow-up. Survival of clinical remission status was estimated using standard Kaplan-Meier analysis for all patients and then compared by IBD type (CD vs UC) using log-rank test. All analyses were performed using R Statistical Software (v4.2.0; R Core Team 2021).

Results

Patient Population

Twenty patients were identified, and all were included in this analysis (Table 1). The median age at baseline was 15 (interquartile range [IQR], 14-16; range, 12-17) years, and median age at diagnosis was 11 (IQR, 10-14) years. Nearly half of the cohort (45%) had CD (6 ileocolonic, 3 colonic; all uncomplicated; 2 with perianal disease); UC (9 pan or extensive disease, 1 left-sided) and IBD-U (1 extensive colitis) comprised the other 55% of the cohort. All but 1 patient (95%) had clinically active disease at UPA start (median [IQR] PCDAI, 20 [17.5-25]; median [IQR] PUCAI, 35 [32.5-47.5]); further, at UPA start, 11 (55%) were on oral steroids, and 5 (25%) were on rectal steroids. Six (30%) patients had concomitant AD. All patients had prior exposure to and failure of a biologic agent; 90% (18/20) had 2 or more biologic therapies fail. Six (30%) were exposed to and had a partial response to tofacitinib.

Table 1.

Demographics and clinical characteristics of adolescents with refractory IBD treated with UPA

Abbreviations: CD, Crohn’s Disease; UC, ulcerative colitis; IBD-U, IBD-unclassified; UPA, upadacitinib; VDZ, vedolizumab; UST, ustekinumab; L1, ileal disease; L2, colonic disease; L3, ileocolonic disease; L4, upper tract disease; p, perianal disease; E2, left-sided disease; E3, extensive disease; E4, pancolonic disease; B1, nonstricturing, nonpenetrating.

UPA Dosing and Strategies

Fifteen (75%) patients were induced with 45 mg daily; 4 of the remaining 5 were started on 30 mg, and a single patient was induced with 15 mg. Upadacitinib was used as monotherapy in a 55% (11/20) and as combination with a biologic in the 9 others (7 with ustekinumab (UST) and 2 with vedolizumab (VDZ). Combination therapy was deescalated in 2 cases; 1 patient stopped UPA after 12 weeks due to insurance issues, continuing on UST alone, and 1 patient stopped UST 16 weeks after starting UPA due to patient preference to continue only 1 therapy. Median (IQR) time on UPA was 271 (176-357) days.

Postinduction Clinical Remission

Steroid-free clinical remission was achieved in 75% (15/20) of patients following induction, with clinical response in 80% (16/20; Figure 1). The SF-CR rate for CD was 44% (4/9), and UC/IBD-U was 100% (11/11). All 3 patients with colonic CD (L2 disease location) were in SF-CR at 12 weeks, while only 1 of the 6 (17%) with ileocolonic CD was in SF-CR at 12 weeks. Patients with UC or colonic CD were significantly more likely to achieve postinduction SF-CR (14/14, 100%) compared with those with ileocolonic CD (1/6, 17%, P < .01). Joint involvement at the start of treatment occurred in 4 patients; 75% of these patients had resolution of joint pain after induction.

Figure 1.

Postinduction and 6-month clinical outcomes in adolescents with IBD on upadacitinib.

6-month Clinical Remission

Two patients stopped therapy with upadacitinib prior to 6 months (1 due to worsening disease symptoms, 1 due to AE). One further patient stopped upadacitinib after induction when in SF-CR due to insurance denial of ongoing coverage. Excluding the patient who stopped due to insurance issues, 74% (14/19) achieved and sustained SF-CR at last follow-up, 56% (5/9) of patients with CD and 90% (9/10) of patients with UC. Survival curves for clinical remission are presented in Figure 2.

Figure 2.

Survival of steroid-free clinical remission on upadacitinib in adolescents with IBD in all comers (left) and by IBD type (right).

Postinduction Objective Outcomes

C-reactive protein was performed a median 16 (IQR, 13-20) weeks on therapy. C-reactive protein was normal in 80% (16/20) of patients. In those with postinduction SF-CR, CRP normality was achieved in all but one.

Fifteen (75%) patients had a baseline and follow-up IUS; follow-up IUS was performed a median of 16 (IQR, 12-20) weeks from UPA start. Two patients had a normal baseline ultrasound; of the 13 with baseline inflammation, 10 (77%) had IUS response. Nearly two-thirds (9/15, 60%) of patients achieved IUS remission.

Influence of UPA strategy on Outcomes

Out of the 11 patients who started on monotherapy, 8 patients (73%) had reached postinduction SF-CR, and 7 (63%) maintained 6-month SF-CR.

Regarding those on combination therapy, in the 7 patients using UPA with UST, all were in SF-CR following induction; one of these stopped UPA due to insurance issues. Five of 6 (83%) then maintained durable SF-CR at 6 months. Of the 2 patients using combination with VDZ, 1 was in postinduction and 6-month SF-CR, and the other patient only achieved SR-CR at 6 months.

Six patients were exposed to tofacitinib prior to their use of UPA; all but one of these patients changed therapy due to an incomplete clinical response to tofacitinib. One patient changed to UPA from tofacitinib due to lack of response of concomitant alopecia; this patient was the single patient in clinical remission at UPA start. Of these patients, 67% (4/6) achieved SF-CR both after induction and at 6 months.

Drug Safety

Over a median of 39 (25-51) weeks of UPA exposure, one SAE and one AE occurred, affecting 2 patients total (10%); none of these events occurred in the first 12 weeks of therapy. One patient was hospitalized and diagnosed with Cytomegalovirus colitis; this was the only infectious event, and it occurred in a patient on 30 mg daily monotherapy after a duration of 26 weeks of UPA total exposure. Notably, this patient had also been on 40 mg of prednisone for 1 week at the time of Cytomegalovirus diagnosis. Upadacitinib was stopped in this case, and the patient ultimately had subtotal colectomy. The AE occurred in a patient on 30 mg daily in combination with ustekinumab who developed hypercholesterolemia, elevations in low-density lipoprotein, and total cholesterol on fasting lipid panel; this was first observed at 32 weeks of therapy, and the patient was managed with serial monitoring alone, with medical treatment deemed unnecessary per endocrinology. No patients experienced a Herpes zoster infection; of note, all patients had completed a 2-vaccination series for varicella zoster.

Discussion

This is the first report of the use of UPA in a cohort of adolescents with IBD. Upadacitinib demonstrated promising effectiveness and safety in a highly refractory cohort at our single tertiary US center. A large majority of patients achieved SF-CR following induction, and in many, this was a durable remission sustained through 6 months. Beyond clinical remission, a subset of patients who underwent baseline and postinduction IUS attained more robust transmural outcomes, with three-quarters demonstrating IUS response and 60% with IUS remission. We also report on the use of UPA after tofacitinib, which we found to be highly effective in a patient population who achieved a prior partial response to tofacitinib; this has similarly been seen in an adult population with IBD.¹⁶ Moreover, patients with UC or colonic CD had a significantly higher rate of postinduction SF-CR compared with those with ileocolonic CD; this has been seen in existing data from both adult real-world cohorts and clinical trials.^4,6,7,17

Our rates of postinduction SF-CR are higher than adult observational studies.^6,17 Diverging from these observational studies, our case series very commonly used combination therapy with a biologic, and it is possible that dual therapy contributed to these improvements in clinical outcomes given recent clinical trials and observational studies showing improvements in outcomes with dual therapy.^18,19 We would venture that the future direction of IBD care is the use of a platform therapy, which is safe, durable, and with low rates of immunogenicity, overlaid with a rapidly acting small molecule therapy, with a broad immunomodulatory effect, that can be used episodically akin to the historical use of steroids. Our most common combination that achieved postinduction SF-CR in all 7 patients in whom it was used was anti-interleukin (IL)-12/23 blockade (UST) and a JAK inhibitor (UPA). We think this is a rational pair, with UST as the platform therapy and UPA as the rapidly acting small molecule, since the IL-23 pathway is distinct from the JAK pathway without significant overlap as can be found with other biologic therapies.²⁰ One of these patients successfully used UPA as a bridge to UST to avoid the use of steroid; this is an illustrative example of this proposed tactic. One patient also successfully deescalated to UPA monotherapy in maintenance; however, deescalation was uncommon at the time of this reporting, as treat-to-target endoscopies are currently being undertaken, around 1 year of follow-up, to guide decisions to deescalate. Importantly, we observed no additional safety signals with our use of combination therapy.

Dual therapy alone may not be the sole explanation for our high rate of postinduction SF-CR, as 73% of those on monotherapy achieved this outcome, which is still high compared with adult studies. It is possible that the induction dose of 45 mg, which was most commonly used, may lead to relatively higher rates of response in adolescents compared with adults. Additionally, it is our practice to target “signature cytokine hubs” based on concomitant immune-mediated inflammatory diseases (ie, the quarter of our patients with concomitant AD, an indication for which UPA is approved in adolescents), and it is possible that this may have enhanced effectiveness by a rough form of precision medicine.²¹

JAKi have been associated with significant AEs, with most data arising from the nonselective inhibitor tofacitinib in an older rheumatoid arthritis population, where increased risks of venothromboembolic disease and malignancy have been observed.²² This risk profile has not been replicated in IBD over 7 years of follow-up,²³ and we did not observe any such events. There has also been a consistent safety signal of an increase in herpes zoster across all JAKi studies.²³ In our cohort, all patients received varicella zoster vaccine, and the risk of herpes zoster is lower in those having received varicella vaccination compared with those who have had natural infection.²⁴ We did not observe any herpes zoster events during the median 39 weeks of therapy, but longer follow-up is needed. While certain AEs of interest with JAKi are dose-dependent, with improvements in safety with reductions in dose,²⁵ we observed only a single infectious event and 1 case of hypercholesterolemia, both on 30 mg daily.

Given the observational, retrospective nature of the study, our study has several limitations, including the small sample size, heterogeneity of dosing, and use of combination therapy, as discussed previously, which is not considered the standard of care. Additionally, our study did not use UPA in patients younger than 12 years old, as there is inherently some limitation in form of the medication for the youngest patients, with no defined dosing regimen, no suspension form, and 3 fixed-dose tablets that cannot be crushed²⁶; this limits our ability to comment on its use in a younger pediatric population. Another unique feature of our population is that adolescent populations may place increased importance on minor side effects, which we did not formally capture with our retrospective design; an important example of this is acne, which is a known side effect of UPA and should be discussed with teenagers during a shared decision-making process.²⁷

In conclusion, we have demonstrated in a real-world case series that UPA is effective in inducing and maintaining remission in adolescents with highly refractory IBD with an acceptable safety profile. Further studies on the efficacy and safety of UPA are needed, including well-powered registration trials.

Conflicts of Interest

S.B.: Support from the Digestive Disease Research Foundation (DDRF) Fellowship Program. M.C.D.: Consulting fees from Abbvie, Amgen, Arena Pharmaceuticals, AstraZeneca, Boehringer -Ingelheim, Celgene, Ferring, Genentech, Gilead, Hoffmann-La Roche, Janssen, Pfizer, Prometheus Biosciences, Takeda, and Target PharmaSolutions. Research funding from Abbvie, Janssen, Pfizer, Prometheus Biosciences, and Takeda. Cofounder and CEO of Trellus Health. M.T.D.: Consulting fees from Neurologica Corp., a subsidiary of Samsung Electronics Co., Ltd. A.K.: No conflict of interest. N.P.: No conflict of interest. D.D.: No conflict of interest. E.A.S.: Grant support from the NIH (K23DK125760-01).