Abstract
With further knowledge of the pathogenesis of inflammatory bowel disease, small oral molecules have become available, including the Janus kinase (JAK) inhibitors. Upadacitinib (UPA) is a selective JAK1 inhibitor and has become the newest drug in this class, with recent approval for the management of moderate-to-severe ulcerative colitis. The large phase III program (including the U-ACHIEVE and U-ACCOMPLISH parallel induction trials and the U-ACHIEVE Maintenance trial) demonstrated superiority over placebo, for all primary and secondary endpoints including key clinical, endoscopic, and histological outcomes utilizing 45 mg orally (po) once daily (OD) during induction and either 30 mg or 15 mg po OD in maintenance. From a safety perspective, UPA has proven to be a safe and well-tolerated medication across immune-mediated diseases with manageable adverse risks such as an increase in herpes zoster. Proper discussion and patient profiling are essential when positioning UPA, considering efficacy and potential risks associated with this highly effective medication.
Introduction
Ulcerative colitis (UC) is a chronic inflammatory bowel disease (IBD) that affects the large intestine, typically involving the rectum and extending proximally. Genetic predisposition in combination with environmental influences are key drivers of its pathogenesis, with changes in the microbiome and alterations in the intestinal barrier playing crucial roles leading to abnormal chronic inflammation. The exact mechanisms and individual processes have not been fully elucidated yet.1–6 UC has a relapsing and remitting behavior. Cardinal symptoms during flares include increased stool frequency, rectal bleeding, abdominal pain, tenesmus, and urgency. This disease has been increasing in incidence and prevalence in most regions of the world, leading to altered quality of life (QoL) and representing a high burden for healthcare systems and society.7–10
The therapeutic goals for UC have evolved over time. Currently, the strategies are focused on achieving symptomatic control in the short term (normalization of bowel movements and the disappearance of rectal bleeding) and on normalizing biomarkers, including fecal calprotectin (FC) and C-reactive protein (CRP), with restoration of growth in children as an intermediate target. Long-term objectives should aim for endoscopic healing, normalization of QoL, and absence of disability, including decreasing the risk of surgery, dysplasia, or colorectal cancer. Histological remission is still considered an aspirational target but has recently been incorporated into the definition of mucosal healing in clinical trials. 11
The availability of advanced therapeutic options for UC has broadened in recent years, consisting of anti-tumor necrosis factor (TNF), including infliximab (IFX), adalimumab, and golimumab; the anti-α4β7 integrin blocker vedolizumab, and more recently ustekinumab (UST), the anti-p40 anti-interleukin (IL) 12/23. With further knowledge of the inflammatory pathways involved in IBD, oral targeted small molecules have been developed, including Janus kinase (JAK) inhibitors of which tofacitinib is approved and ozanimod, a sphingosine-1-phosphate receptor modulator.12–19 Despite these options, the rate of non-responders and secondary loss of response is still high in the clinical setting. 20 This gap has led to the creation of the ‘second-generation’ selective JAK inhibitors, leading to the approval of filgotinib (FIL) in Europe and Japan 21 and, more recently, upadacitinib (UPA), a selective and reversible JAK 1 inhibitor in the United States, Australia, Canada, and Europe.22,23 In this review, we discuss the available literature regarding UPA, the data supporting its selectivity, and the clinical development programs outlining its efficacy and safety in moderate-to-severe active UC.
Methodology
Relevant publications on Medline/PubMed were identified up to 30 September 2022, with the terms: ‘upadacitinib’, ‘ulcerative colitis’, ‘JAK inhibitor’, ‘safety’. The authors independently evaluated the abstracts to identify eligible studies, and discrepancies were discussed among the co-authors.
The JAK-signal transduction and activation of transcription pathway and role in IBD
The description in 1992 of the JAK-signal transduction and activation of transcription (JAK-STAT) pathway has provided a better understanding of intracellular signaling, its importance, and its role in hematopoiesis, immunity, inflammation, growth, neural development, and metabolism.24,25
The JAK-STAT pathway involves the transmission of signals from outside the cell to inside the cell, beginning with the engagement of transmembrane receptors and culminating in signal transduction and transcription at the level of the nucleus. 25 Cytokines utilizing the JAK-STAT can be classified according to the kind of receptors they bind to; and type I and II cytokine receptors are dimeric structures composed of two or more different subunits which are JAK dependent. Type I and type II cytokines include, among others: IL-2, IL-3, IL-4, IL-6, IL-12, IL-23, interferons, and IL-10-related cytokines.25–27 There are four JAKs: JAK1, JAK2, JAK3, and TYK2, which are tyrosine kinases working in pairs with the potential to phosphorylate cytokine receptors, other JAKs, or themselves, as well as cytoplasmic STAT proteins. There are several combinations of JAKs associated with different cytokine receptors, which, in turn, modulate a variety of physiologic processes (see Table 1).25,28,29
Cytokines and signal effects using JAK-STAT intracellular signaling.
EPO, erythropoietin; G-CSF, granulocyte colony-stimulating factor; GH, growth hormone; GM-CSF, granulocyte–macrophage colony-stimulating factor; IFN, interferon; IL, interleukin; JAK, Janus kinase; LIF, leukemia inhibitory factor; OSM, oncostatin; STAT, signal transducer and activator of transcription; TPO, thrombopoietin activator of transcription; TYK2, tyrosine kinase 2.
The STAT group comprises seven proteins: STAT1, STAT2, STAT3, STAT4, STAT5A, STAT5B, and STAT6. With phosphorylation, STAT proteins have a structural change that allows them to form homodimers or heterodimers. This conformation translocates to the nucleus, binds to the DNA, and regulates genetic transcription (Figure 1). The STAT signal will vary according to the cytokine receptor activated24,25,28–30,32.
JAK/STAT pathway: activation and inhibition.
JAK-STAT signaling plays a vital role in intestinal mucosal immunology and hemostasis, with roles in epithelial proliferation, differentiation, and apoptosis (see Table 1). 27 Genetic association studies have demonstrated that polymorphisms in the receptors, JAKs, and STAT genes are related to a higher risk of developing IBD. 30 There is increased intracellular activity in this pathway in IBD patients because several of the cytokines in the innate and adaptative immunity involved in the pathogenesis of IBD use it as intracellular signaling (see Table 2). The JAK-STAT signals can activate B cells and differentiate and expand naïve T helper (Th0) into Th1, Th2, Th9, and Th17 effector lymphocytes. At the same time, these cells use the JAK-STAT pathway to modulate other immune cells (e.g. macrophages, dendritic cells, regulatory T cells, and CD8+ cytotoxic cells).27,29,33 The knowledge of these concepts has led to the discovery of a novel series of oral drugs with a new mechanism of action in treating IBD and other immune-mediated conditions. 34
AQP, aquaporin; CBC-ISCs, crypt-based columnar intestinal stem cells; GATA3, GATA-binding protein 3; IECs, intraepithelial cells; IFN, interferon; IL, interleukin; ISCs, intestinal steam cells; JAK, Janus kinase; LPS, lipopolysaccharides; RORγt, retinoic acid receptor-related orphan receptor γ; SOCS, suppressors of cytokine signaling; STAT, signal transducer and activator of transcription.
The rationale for JAK1 selectivity
There are a wide variety of cytokines JAK-STAT signaling pathway; importantly, JAK1 is part of the specific JAK combinations participating as intracellular effectors for over three-quarters of the main pro-inflammatory cytokines involved in the pathogenesis of IBD (see Table 3).32–34 For example, it has been demonstrated that IL-6, a potent pro-inflammatory cytokine in IBD, activates JAK1 as a dominant kinase in in vivo models, with little relevance to the other JAKs.37,38 This JAK1 activation increases the STAT3 phosphorylation in a higher proportion than the rest of the STAT family, with an active role in T-cell proliferation and survival.32,39 Tofacitinib (TOFA) is a pan-JAK inhibitor, initially approved by the FDA in 2012 for rheumatoid arthritis (RA) 40 and in 2018 for UC 41 as the first advanced oral therapy in the field. However, concerns about long-term safety have arisen with pan-JAK inhibition; including herpes zoster (HZ) reactivation, hypercholesterolemia, venous thromboembolic (VTE) episodes,34,42–44 and more recently, mortality, major adverse cardiovascular events (MACE), and neoplasia in high-risk patient populations with RA. Many of these adverse events (AEs) appear to be dose related with pan-JAKs. 45 Based on this knowledge about the role of JAK1 in the intracellular signals in IBD as a common ground intracellular signal for several pro-inflammatory cytokines, newer selective and reversible JAK1 inhibitors, such as UPA, have been developed. JAK 1 selectivity promises to exhibit higher efficacy while preserving safety and thereby increasing the therapeutic window by reducing some dose-related AEs described with TOFA. 46 Also, given the relevance of the pro-inflammatory cytokines in the differentiation of T cells into different pathogenic Th responses and the suppression of Treg cellular functions, it is important to use a therapy that can selectively inhibit as much as possible the pathogenic T cells, helping to restore the Treg balance and control inflammation. 29
Cytokines involved in IBD pathogenesis with JAK-STAT intracellular signaling.
IBD, inflammatory bowel disease; IL, interleukin; IFN, interferon; JAK, Janus kinase; STAT, signal transducer and activator of transcription.
UPA preclinical and pharmacological data
UPA is an oral, once-a-day, small molecule selective JAK1 inhibitor. It has been designed specifically to be a highly selective competitive inhibitor of the adenosine triphosphate binding pocket in the JAK1.46,47 UPA has been proven in cellular assays to have a multi-fold selectivity for JAK1 versus JAK2 of 42, JAK1 versus JAK3, and JAK1 versus TYK2 of 133 and 194, respectively. 46 UPA has a half-life of 4 h, and it is mainly metabolized in the liver by the cytochrome P450 (CYP3A4) and, to a smaller extent, by the CYP2D6. Potent CYP3A4 inhibitors can increase UPA exposure by 75%, while inducers of the cytochrome will reduce its half-life by half. 48 Phase I studies suggest drug doses do not need to be adjusted with race or age, 48 mild-to-moderate renal impairment [estimated glomerular filtration rate (eGFR) > 30 mL/min/1.73 m2], 49 or mild and moderate liver dysfunction (Child-Pug A and B). 50 In rat adjuvant-induced arthritis (AIA) models, UPA proved more effective than TOFA at comparable exposure. To achieve similar results, higher doses of TOFA were required to achieve the same response as UPA. 46 Both drugs were also assessed in the AIA models for JAK2 and JAK3 inhibition, resulting in a lesser effect on reticulocytes and natural killer cell counts for UPA at comparable efficacy levels with tofacitinib.46,47
UPA clinical trials in UC
Phase II trial in UC
A phase IIb dose-ranging induction study (substudy 1) was part of the U-ACHIEVE program to assess the efficacy and safety of UPA in moderate-to-severe UC. This program was a multicenter, randomized, double-blind, placebo-controlled trial. In part 1 of the study, 250 UC patients were randomized in 1:1:1:1 to receive either placebo or four doses of UPA; 7.5, 15, 30, or 45 mg once daily (OD) for 8 weeks. In part 2 of the study, 132 patients were added to the UPA 30 mg and 45 mg groups to provide a sufficient number of clinical responders for the subsequent phase III maintenance trial. 51 Eligible patients had moderate-to-severe active disease, defined as an adapted Mayo score (AMS) of 5–9 points with a centrally read Mayo endoscopic subscore (MESS) ⩾ 2. Patients had to have a history of inadequate response to corticosteroids, immunosuppressive agents, and/or biologics. The use of oral aminosalicylates, methotrexate, and oral corticosteroids (⩽30 mg/day of prednisone or equivalent) was permitted and unchanged during the study. Concomitant thiopurines, biologics, intravenous steroids, previous exposure to TOFA, or other JAK inhibitors were not allowed. 51 The primary endpoint for this study was clinical remission at week 8 (AMS with a stool frequency subscore (SFS) of ⩽1, rectal bleeding subscore (RBS) of 0, and MESS ⩽ 1) (see Table 4 for further definitions). This outcome was achieved in 0% of the patients in the placebo arm, 8.5% (p = 0.052) with UPA 7.5 mg, 14.3% (p = 0.013) with 15 mg, 13.5% (p = 0.011) with 30 mg and 19.6% (p = 0.002) with 45 mg. 51
All the endoscopy was centrally read.
AMS, adapted Mayo score; CSF, corticosteroid-free; FMS, full Mayo score; HEMI, histological-endoscopic mucosal improvement; MESS, Mayo endoscopic subscore; PAMS, partial adapted Mayo score; RBS, rectal bleeding subscore.
A number of secondary endpoints were assessed, endoscopic improvement (MESS ⩽ 1 without friability) at week 8 was 2.2% with placebo, 14.9% (p = 0.033), 30.6% (p < 0.01), 26.9% (p < 0.01), and 35.7% (p < 0.01) for UPA 7.5, 15, 30, and 45 mg, respectively. Clinical remission according to the full Mayo score (FMS) at week 8 was in patients receiving UPA 15 mg (10.2%, p = 0.027), 30 mg (11.5%, p = 0.016), 40 mg (19.6%, p = 0.001) versus 0% in the placebo group. Endoscopic remission (MESS of 0) was reported in 9.6% (p = 0.015) for UPA 30 mg and 17.9% (p = 0.004)with UPA 45 mg. 51 There was a reduction in the reported bowel urgency and abdominal pain at week 8. In subsequent analyses, symptom improvement was reported as early as week 2, with UPA patients reaching statistical significance, representing a rapid control of the symptoms that affect the QoL. 53
Clinical response according to the AMS (decrease from baseline in the AMS of ⩾2 points and ⩾30% from baseline, plus a decrease in RBS of ⩾1 or an absolute RBS of ⩽1) was documented at 8 weeks in 29.8% (p = 0.046) of the patients in the 7.5 mg arm, 44.9% (p < 0.001) with 15 mg, 44.2% (p < 0.001) with 30 mg, and 50.0% (p < 0.001) in the 45 mg group compared to 13.0% with placebo. Histologic improvement (any decrease from baseline in Geboes Score) at 8 weeks was reported in 31.9% (p = 0.003), 51.0% (p < 0.001), 44.2% (p < 0.001), and 48.2% (p < 0.001) within the same UPA groups, respectively, and only in 6.5% in the placebo arm. 51 Regarding high-sensitivity CRP and FC, there was a statistically significant reduction from baseline at 8 weeks within all the groups that received active therapy compared with placebo. 51 Interestingly, when the groups were stratified accordingly to previous exposure or not to biological treatments, a positive response was observed in all groups with active medication; the highest response rates were seen with UPA 45 mg and were numerically higher in the bio-naïve patients. 51 This dose-ranging study determined that 45 mg OD was likely the optimal dose for induction and was brought forward into the phase III program.51,52
Phase III trials in UC
The phase III program consisted of two comparable induction trials [U-ACHIEVE; substudy 2 (UC1, n = 474 patients) and U-ACCOMPLISH (UC2, n = 522 patients)], and a maintenance study [U-ACHIEVE Maintenance substudy 3 (UC3, n = 451 patients)]. The inclusion and exclusion criteria for the induction trials were the same as for the phase IIb substudy 1. Patients were randomized in a 2:1 manner to receive either UPA 45 mg OD or placebo for 8 weeks. The primary endpoint at week 8 was clinical remission, defined as SFS less than equal to 1, RBS of 0, and a MESS of 0 or 1 with no friability. At week 8, responders according to the AMS (including patients from substudy 1 that responded to UPA 45 mg) were re-randomized in the maintenance study (U-ACHIEVE maintenance, substudy 3) to receive UPA 15 mg OD, UPA 30 mg OD, or withdrawn to placebo for an additional 52 weeks. 52 In U-ACHIEVE, clinical remission at 8 weeks was achieved by 26.1% in the UPA group versus only 4.8% with placebo (p < 0.0001). Subgroup analysis by previous biologic failure demonstrated clinical remission in 17.9% with UPA 45 mg daily versus 0.4% in the placebo group, compared to remission of 35.2% versus 9.2% for UPA and placebo, respectively, in the bio-naïve subgroup. 52 In U-ACCOMPLISH, similar outcomes were demonstrated, with 33.5% of patients achieving clinical remission at week 8 in the UPA group and 4.1% with placebo (p < 0.0001). Similar to UC1, UPA 45 mg was superior to placebo in patients with previous biologic failure (29.6% versus 2.4%) and without biologic failure (37.5% versus 5.9%). 52 Therefore, both induction studies met the primary endpoint of clinical remission with large deltas over placebo despite recruiting patients with a high inflammatory burden and substantial previous exposure to multiple therapies, including all classes of available biologics.
The key secondary outcomes at 8 weeks were all met, including clinical response, rapid clinical response/remission, endoscopic response and remission, and histological-endoscopic mucosal improvement (HEMI). Highlighting these endpoints was a very high overall response rate of nearly 75%. In U-ACHIEVE, clinical response by AMS was 73% with UPA 45 mg daily and 27% in the placebo group (p < 0.0001), with clinical response occurring rapidly by partial adapted Mayo score (PAMS) at week 2 of 60% in the UPA group versus 27% with placebo (p < 0.0001). A post-hoc analysis revealed separations from placebo in stool frequency, abdominal pain, rectal bleeding, and urgency within 1–3 days of initiating the 45 mg dose. 54 The rate of clinical remission per FMS at 8 weeks was 23.0% versus 2.8% in the placebo group (p < 0.001). 52 Endoscopic improvement was demonstrated in 36%, endoscopic remission of 14%, and HEMI (MESS ⩽ 1 without friability and Geboes score ⩽ 3.1) was achieved in 30% of the UPA patients, with a histological improvement of 55% and mucosal healing (MESS of 0 and Geboes score < 2) of 11%, all outcomes superior to placebo, p < 0.0001. The disappearance of bowel urgency happened in 48% and disappearance of abdominal pain in 47% of the UPA patients (p < 0.0001). 52 The U-ACCOMPLISH trial reported similar findings in the secondary outcomes at week 8. In the UPA 45 mg group, clinical response by AMS was 74%, and clinical response by PAMS at week 2 was 63% compared to placebo. The clinical remission per FMS at 8 weeks was 27.3% versus 2.4% with placebo. The endoscopic improvement rate was 44%, a 35.1% difference compared to placebo, with a reported endoscopic remission of 18% versus placebo, p < 0.0001 for all comparisons. 52 HEMI was achieved significantly in 36% of the patients with UPA, and mucosal healing was reached in 11.3% more than in placebo, p < 0.0001. Regarding the disappearance of bowel urgency and abdominal pain, the rates were 27.1% and 29.1%, respectively, for the UPA group; also, the same group had an IBD questionnaire (IBDQ) reduction of 52.2 points from baseline. 52
Biomarkers were assessed during the induction trial in these pivotal studies. There was a significant drop in FC, with a more substantial proportion of patients in the UPA groups achieving values <150 mg/kg at week 2 (29.4% versus 4.8% in the placebo group for UC1; and 29.8% versus 5.4% with placebo in UC2) and week 8 (44.4% and 40.4% versus 7.9% and 7.7% in both induction studies versus placebo, respectively). High-sensitivity CRP concentrations were reduced in a greater proportion of patients with UPA 45 mg daily compared to placebo in both induction studies. This trend was observed as early as week 2 of therapy (first monitoring visit). 52
The induction studies design allowed for a period of extended induction in non-responders at week 8. In total, 125 patients with active medication failed to achieve clinical response at week 8 and completed another 8-week open-label induction with 45 mg. Responders per AMS to this extended induction were randomly assigned to receive either 15 mg or 30 mg UPA daily for 52 weeks. 52 Clinical response was seen in 48.3% and clinical remission in 5.6% of the extended induction patients at week 16. At week 52, UPA 30 mg achieved outcomes in a higher proportion than UPA 15 mg; clinical remission was documented in 33.3% versus 19.0%, clinical response in 66.7% versus 35.7%, and endoscopic improvement in 37.5% versus 23.8%, respectively. There were no new safety signals with an extended induction strategy. 55
In the U-ACHIEVE maintenance, the primary outcome was clinical remission at week 52; a two-Fisher’s exact test at a 0.025 significant level with multiplicity of adjustment was used for this analysis. This endpoint was reached in 42% of the UPA 15 mg group and 52% with UPA 30 mg, compared with 12% of the patients with placebo, p < 0.0001. 52 In stratified subgroup analyses for the primary endpoint, UPA 15 mg and 30 mg were superior to placebo. Clinical remission rates difference adjusted for placebo in UC patients with a baseline AMS ⩽ 7 were 37.8% and 46.2% for UPA 15 mg and 30 mg, respectively. At the same time, patients with an AMS > 7 at enrollment achieved an adjusted difference with placebo of 19% (UPA 15 mg) and 32% (UPA 30 mg). The remission rate at 52 weeks for patients with disease duration ⩽6.1 years was 40.9% for UPA 15 mg and 49.6% for UPA 30 mg, while in patients with longer disease duration was 43.7% and 53.8% for 15 mg and 30 mg, respectively. In general, UPA 30 mg presented a numerical superiority over UPA 15 mg within the subgroups (Table 5). 52 Secondary endpoints in the maintenance study demonstrated maintenance of clinical remission per AMS of 57% (UPA 15 mg), 68% (UPA 30 mg), versus 22% in the placebo group (p < 0.0001). The adjusted treatment difference with placebo for corticosteroid-free (CSF) clinical remission was 35.4% and 45.1% (p < 0.0001) for UPA 15 mg and 30 mg, respectively. Maintenance of clinical response by AMS was 19% in the placebo group, 63% in the UPA 15 mg group, and 77% UPA 30 mg arm (p < 0.0001). 52 Endoscopic remission rates were 26%, 24%, and 6% for UPA 30 mg, 15 mg, and placebo, respectively, while endoscopic improvement was demonstrated in 62%, 49%, and 14% of the same groups. Regarding HEMI, 50% of the 30 mg group and 35% of the 15 mg group reached this goal compared to only 12% with placebo. Mucosal healing was achieved in 19%, 18%, and 5% of the same groups (p < 0.0001 for all analyses). 52 QoL was assessed in maintenance as well. The reduction from baseline in the IBDQ total score was 58.9 and 49 points in the UPA 30 mg and 15 mg groups, respectively, and just 17.9 in the placebo group. The adjusted treatment difference with placebo for the absence of urgency and abdominal pain at 52 weeks was 45.1% and 33.7% in UPA 30 mg, 38.7%, and 24.3% in UPA 15 mg, respectively (all p < 0.0001). 52
Subgroup analyses for the primary endpoint in the U-ACHIEVE maintenance study at week 52 (intention-to-treat population).
AMS, adapted Mayo score; CS, corticosteroids; FMS, full Mayo score; 5-ASA, aminosalicylic acid.
The subgroup analysis of the primary and secondary endpoints at week 52 in patients with and without previous biologic failures revealed overall superiority to placebo. However, the data are consistent with the numerical superiority of UPA 30 mg over 15 mg, especially in patients with a higher disease burden. 52 Clinical remission was 54%, 43.9%, and 17.6% for UPA 30 mg, 15 mg, and placebo with no previous biologic failure. In contrast, it was achieved in 49.1%, 40.5%, and 7.5% for previous biologic failures in the same group order. CSF clinical remission adjusted treatment difference with placebo was 59.4% (UPA 30 mg) and 57% (UPA 15 mg) for previous biologic failures and 37.2% and 21.3% in the same groups for no previous biologic failures. 52 At 52 weeks, the incidence of UC-related hospitalizations was reduced by 6.36 (nominal p = 0.023) with UPA 15 mg and 5.77 (nominal p = 0.044) with UPA 30 mg. 52
Extraintestinal manifestations (EIM) were present in 25.0% of the UPA induction groups and 26.5% of the placebo patients. Anemia, peripheral arthropathy, and axial arthropathy were the most commonly reported. After 8 weeks of induction with UPA 45 mg, the resolution of EIM was numerically higher than with placebo (40.0% versus 33.3%). At week 52, the resolution was 65.9% with UPA 30 mg versus 24.3% in the placebo group (p < 0.001) and 41.7% with UPA 15 mg (numerical difference). 56
Comparative efficacy of UPA with other advanced therapies for UC
Head-to-head trials continue to be the gold standard for direct comparisons of therapies in IBD because they are designed and powered for this purpose. 57 Currently, there are no trials studying weighing directly the efficacy of UPA against other advanced therapies available for UC; however, indirect comparisons are inevitable, and several network meta-analyses (NMA) have been published recently assessing the efficacy and safety of advanced therapies for moderate-to-severe UC. Burr et al. published the first systematic review and NMA identified 28 UC trials with reported clinical remission between weeks 6 and 14, including 12,504 patients from different clinical studies with anti-TNF therapies, anti-integrins, UST, TOFA, and the JAK-1 selective FIL. Based on the inability to achieve clinical remission, UPA 45 mg daily was ranked as the most effective therapy versus placebo, RR 0.78 (95% CI: 0.72–0.85), P-score 0.98. This was followed by IFX 5 mg/kg and IFX 10 mg/kg, RR 0.78 (95% CI: 0.72–0.84), P-score 0.94 and RR 0.80 (95% CI: 0.72–0.89), P-score 0.84, respectively. 58 Interestingly, TOFA 10 mg twice a day, ranked in fourth position (95% CI: 0.80–0.93; RR: 0.86, P-score 0.64), while FIL 200 mg [RR 0.0.91 (0.84–0.99), P-score 0.37] and FIL 100 mg [RR 0.97 (0.90–1.05), P-score 0.13] were placed in 13th and last positions, respectively. In a stratified analysis, UPA 45 mg ranked first, RR 0.30 (0.23–0.40), P-score 1.00, and RR 0.78 (0.72–0.85), P-score 0.99 for naïve and previous exposure to anti-TNF therapy, correspondingly. 58 Similarly, the analysis for failure to achieve endoscopic improvement, positioned IFX 10 mg/kg (off-label dose) first in the ranking [RR 0.61 (0.51–0.72), P-score 0.97], followed by UPA 45 mg [0.65 (0.61–0.78), P-score 0.93]. In the sub-analysis by previous exposure or not to anti-TNF therapies, UPA 45 mg was ranked first in both groups. 58
A second NMA using Bayesian methodology assessed the safety and efficacy of advanced therapies for induction (6–10 weeks) and maintenance (44–54 weeks post-induction) in moderately to severely active UC. In all, 23 randomized clinical trials were included in the analysis, considering all the currently approved advanced therapies for UC. 59 This NMA converted the response rates in the individual trials to approximate a treat-through design. In the bio-naïve population, over all the analyzed therapies, UPA 45 mg ranked first in achieving clinical remission during induction [OR 9.6 versus placebo, number needed to treat (NNT) 2.5, surface under the cumulative ranking curve (SUCRA) 97%). It also ranked first in achieving clinical response and endoscopic improvement during the same period. Similarly, UPA 30 mg was positioned on top as maintenance therapy for clinical remission (OR versus placebo 4.2, NNT 3.3, SUCRA 72%). It also ranked first for clinical response and endoscopic improvement as maintenance therapy. 59 Similar results were seen in the analysis with the bio-exposed population; UPA 45 mg during the induction phase and UPA 30 mg as maintenance therapy were placed first in the ranking of advanced treatments to induce clinical response, clinical remission, and endoscopic improvement. 59 The maintenance of efficacy rates was adjusted by the likelihood of induction response for all the medications in an intention-to-treat analysis. For both bio-naïve and bio-exposed patients, UPA showed the highest consistency as the most effective strategy among all the drugs included. 59
Finally, a systematic review and NMA by Lasa and colleagues collected phase III placebo-control or head-to-head randomized control trials (RCT) for UC with 10,061 patients. For induction of clinical remission, UPA was superior to all other therapeutic strategies with moderate-to-high confidence and ranked the highest among all drugs (SUCRA 0.996). The scenario was presented similarly in the analysis for endoscopic improvement, with UPA showing superiority among the different medications with moderate-to-high confidence and placed in first place in the induction of endoscopic improvement (SUCRA 0.999). For maintenance of clinical remission, UPA was significantly superior to placebo, but no difference between UPA and the other therapies was demonstrated. 60
The data generated with the NMAs should be interpreted cautiously; even though these studies only included phase III and head-to-head trials, they have limitations, including different designs (treat-straight-through versus randomized responders). Also, some of the trials for newer medications may include more strict definitions for clinical and endoscopic remission, central endoscopic reading, and in general, more refractory UC patients with previous failures to several therapies.58,60 The data of the NMAs are restricted with a stringent assumption of transitivity, also by the individual limitations of every specific trial included in the network. Ranking the data usually considers only the effect estimate, meaning that some therapy with smaller and lower quality supporting data can still be ranked high, or highly ranked medications can have just a modest or not even clinically relevant treatment effect.61,62
Safety considerations for UPA in UC
The FDA has approved UPA for moderate-to-severe UC patients who have had an inadequate response or intolerance to one or more anti-TNFs. As per the label, it should not be used with other JAK inhibitors, biological therapies, or immunosuppressants like azathioprine and cyclosporine. It could be prescribed concomitantly with methotrexate. 63 For UC in patients with liver impairment (Child-Pugh A or B) or diminished renal function (eGFR 15 to <30 mL/min/1.73 m2), the recommended induction dose is 30 mg daily for 8 weeks and maintenance of 15 mg daily. 63 The treatment is not indicated during pregnancy and breastfeeding. Also, it should be interrupted with an absolute neutrophil count of <1000 cells/mm3, a total lymphocyte count of <500 cells/mm3, hemoglobin <8 g/dL, and if there is suspected drug-induced liver injury. 63 There is a warning box regarding an increased risk of severe infections with UPA. This risk has been higher in patients over 65 years old. 63 Also, the label is consistent with language for all JAK inhibitors based on the result of the ORAL SURVEILLANCE study, a large event-based phase IV study comparing TOFA versus anti-TNF in a high-risk RA population, 45 referencing an increased risk of mortality, malignancy, MACE, and VTEs; this recommendation for UC patients is based on this extrapolated data.45,63 However, the EMEA has approved the therapy for patients with moderately to severely active UC who have failed mesalamine, corticosteroids, or immunosuppressants. There is no restriction on having failed anti-TNF or other advanced therapies.
In the phase IIb U-ACHIEVE trial, the rate of AEs was numerically higher in the placebo group in comparison to the rest of the UPA groups, including the rate of any serious AE (SAE) (10.9% with placebo, 0% UPA 7.5 mg, 4.1% UPA 15 mg, 5.8% UPA 30 mg, and 5.4% UPA 45 mg). 51 There was only one case of HZ reported with UPA in the 45 mg group. A blinded independent committee assessed cardiovascular, embolic, and thrombotic events. Only one patient with UPA 45 mg had a documented pulmonary embolism and deep VTE 26 days after the medication discontinuation due to worsening of UC. 51 In the phase III induction studies, nasopharyngitis, creatinine kinase (CPK) elevation, and acne were the most frequent AEs. In both studies, the rate of SAE and AE leading to discontinuation was lower for UPA compared to placebo (3% versus 6% and 2% versus 9% in UC1) and (3% versus 5%, and 2% versus 5% in UC2), respectively). 52 Three cases of HZ were documented with UPA 45 mg, while one gastrointestinal perforation and one VTE episode were presented on the placebo arm. 52
In the maintenance trial, AEs were reported more frequently in the active medication arms, mainly nasopharyngitis and CPK elevation, none leading to therapy discontinuation. Worsening of UC was seen in 30% of the placebo patients, in contrast to 13% and 7% with UPA 15 mg and 30 mg, correspondingly. The rate of SAE and AE leading to discontinuation was higher in the placebo arm (13% and 11% for placebo, 7% and 4% for UPA 15 mg, 6% and 6% for UPA 30 mg, respectively). 52
The serious infection rate was 4% for the placebo group and 3% for both UPA patients. There were cases of HZ reported in 4% of both UPA arms, and no events were documented with placebo. Regarding malignancies, two patients developed breast cancer, one with placebo and one with UPA 15 mg. One case of colonic cancer, two non-melanoma skin cancer (NMSC), and one case of prostate cancer were reported with UPA 30 mg. 52 Only one MACE was documented in the placebo group, and no events in the UPA arms. Two cases of VTE were reported with UPA 30 mg; however, both events were unrelated to the study drug (as per independent committee determination), and no documented VTE presented with UPA 15 mg and placebo. These trials did not report any deaths. 52 Long-term changes in the lipid profile in the UPA-exposed patients revealed an increase in the total cholesterol concentrations; however, the proportion of low-density lipoprotein cholesterol and high-density lipoprotein cholesterol remained stable and was not significantly different from placebo. Reported cases of anemia were higher in the placebo group, and neutropenia was higher with UPA 30 mg. Nevertheless, these findings were not severe and did not lead to therapy discontinuation. 52
A meta-analysis of IBD patients exposed to different JAK inhibitors, including patients treated with UPA, revealed that there was not an increment of any AE [RR 1.02 (95% CI: 0.97–1.09), I 2 = 0.0%, p = 0.808), in the subgroup analysis, the risk of AE was not increased in CD or UC. 64 There was not an increased risk of SAE with JAK inhibitors in IBD [RR 0.82 (95% CI: 0.58–1.16), I 2 = 20.1%), and the subgroup analysis revealed a protective effect for SAE in UC patients [RR 0.6 (95% CI: 0.42–0.91), I 2 = 0.0%]. The meta-analysis did reveal an increased risk of infections, particularly HZ, in IBD patients [RR 1.40 (95% CI: 1.18–1.67), I 2 = 0%, p < 0.001). 64 Another systematic review and meta-analysis explored the safety of JAK inhibitors in IBD patients and other immune-mediated diseases, including RA, psoriatic arthritis, atopic dermatitis, and ankylosing spondylitis. It included 82 studies, including different UPA studies, with 66,159 patients and demonstrated an increased risk of HZ in patients exposed to JAK inhibitors [RR 1.57 (1.04–2.37), I 2 = 0.0%, p = 0.034 for overall effect). 65 There was no increment in the risk of AE, SAE, serious infections, NMSC, other malignancies, MACE, VTEs, or mortality in controlled studies. 65 A Bayesian NMA of 23 RCT proved no significant differences in SAE or serious infections for any advanced UC therapies included compared to placebo. The discontinuation due to AE was lower for UPA 45 mg matched to placebo in the induction phase (OR: 0.2, SUCRA: 90%). This risk was lower for UST and vedolizumab during the maintenance phase (OR: 0.2, SUCRA: 89%, and OR: 0.4, SUCRA: 72%, respectively). 59 Regarding safety data from the meta-analysis by Oliveira and colleagues, no differences were observed between all the therapeutic options regarding AE and SAE. UPA ranked the highest for AE, SUCRA 0.843, while ozanimod and placebo were ranked in the highest positions for SAE, SUCRA 0.831 and 0.783, respectively. 60
Expert commentary
The therapeutic armamentarium of available therapies to treat moderate-to-severe UC has expanded in recent years. Advanced treatments were initially in the form of monoclonal antibodies, but more recently, small oral molecules such as JAK inhibitors and S1P modulators hold significant promise. As an oral selective and reversible JAK1 inhibitor, UPA has proven effective in the induction and maintenance of clinical remission in naïve and previous biologic failure UC patients.51,52 It is a practical once-a-day oral extended-release tablet resulting in rapid symptomatic control, which occurs in days in a large proportion of patients.52–54 Given the rapidity of the onset of effect combined with the fact that there is no need for injection training or scheduling of intravenous visits makes this a very attractive ‘out the door’ option. This characteristic can be used to avoid the use of corticosteroids during induction, especially in patients who tolerate corticosteroids poorly have relative contraindications, or are refractory to corticosteroids. 66
The central question is where does UPA fit in the therapeutic cascade. In the absence of head-to-head trials with UPA, recent systematic reviews and NMA with indirect comparisons have suggested that UPA is the most effective therapy ever studied in moderate-to-severe UC during induction and maintenance, regardless of previous history of biologic exposure. Within the JAK class, UPA appears to be more effective than tofacitinib and the other JAK1 selective inhibitor, filgotinib.58–60 Whether this can be attributed to selectivity or the use of a more appropriate dose is debatable. The label in different jurisdictions may limit where it is used. In the United States, where patients need to have been exposed to an anti-TNF, the data suggest that it should be the preferred agent in anti-TNF-exposed patients. 23 In Europe, 22 Australia, 67 it can be positioned as a first-line or second-line therapy. The seminal information from the NMAs provides a rationale to consider UPA early in the therapeutic algorithm and should be offered as first-line therapy after careful and thoughtful discussion with the patient. In the more refractory patients, especially those having failed one or more mechanisms of action, it once again should be top of mind as an effective therapeutic strategy in the authors’ opinion, given the data at hand, it is a more sensible choice than an anti-integrin or anti-IL-12/23, especially in anti-TNF-exposed patients.
Current evidence suggests that considering an extended induction up to 16 weeks when needed expands the proportion of patients who may benefit from UPA. Healthcare providers need to individualize the need and rationale of extended induction therapy based on the label and the patient in front of them. UPA 30 mg daily as maintenance therapy appears to be more effective than 15 mg OD across most endpoints with a delta of anywhere between 8% and 13%. Over 52 weeks of maintenance, there appears to be no difference in toxicity between the doses. The 30 mg dose in maintenance should be particularly considered in patients with pancolitis, previous biologic failure, and EIMs.52,56 New evidence has shown the efficacy of UPA in Crohn’s disease phase III trials with no new safety signals, opening the door for further and more practical use of this highly effective drug. 68
As opposed to anti-integrin or even anti-IL 12/23, UPA has shown a positive impact on controlling EIM in UC patients. 56 There is current evidence of efficacy in other immune-mediated conditions, including RA, psoriatic arthritis, ankylosing spondylitis, and atopic dermatitis. It becomes an attractive option for patients with associated EIM.69–73 A discussion should be held with other specialties regarding the need for induction, extended induction, and probable higher doses for maintenance in UC according to the previous history and disease activity in every case.
Considering the evidence from the ORAL Surveillance study of TOFA versus anti-TNF in RA 45 and the warning issued for all JAK inhibitors by the FDA, it is vital to understand the potential risks in IBD. The ORAL surveillance population was a highly selected population chosen to ensure that the events of interest, namely MACE and malignancy, would occur as it was an event-based study. Therefore, the included population in the trial was >50 years old with known cardiovascular risk factors. The study was not powered to compare specific TOFA doses with anti-TNF for any outcomes.45,74 Even in this population, the differences in MACE and malignancy events were largely driven by those older than age 65, smokers, and those with previous cardiac events. In this high-risk population, a thorough understanding of the rarity events is imperative to discuss with patients properly. The IBD population is significantly different than this high-risk patient population. This difference is highlighted in a recent publication looking at rates of AEs of special interest with TOFA over 7.8 years of exposure. 75 This observation showed no detectable increase in the risk of solid organ malignancy, MACE, or VTE. There was a signal once again for HZ infection. Therefore, tailoring IBD therapies involves understanding the individual patient’s risk profile. A shared decision should be made with each patient discussing the benefits and potential risks, considering every option.
When considering UPA safety, it has proven to be a well-tolerated and safe medication, with no significant differences with placebo within the different doses used.51,52 When looking across indications where UPA has been studied, there does not seem to be an increased risk of AE and SAE aside from the HZ signal, even when compared to adalimumab. 76 Considering the safety concerns around all JAK inhibitors; in IBD there has not been a higher risk of AE and SAE, including MACE or VTE, in clinical trials and meta-analyses. Infections can be reduced with proper screening for active and latent infectious diseases before starting therapy and completing vaccination schemes. Particularly for HZ, now with the availability of a recombinant shingles vaccine, this risk can be lowered considerably.77,78 Follow-up is recommended in all patients starting UPA, documenting baseline lipid profiles, blood cells count, hepatic enzymes [liver function test (LFT)], CPK, and serum creatinine. Lipids should be reassessed approximately 8–12 weeks after starting therapy. The authors recommend hematological parameters and LFT assessment at 8–12 weeks after UPA initiation and every 3 months for the first year thereafter.79,80
Real-world evidence and head-to-head data are needed to further validate the clinical efficacy and long-term safety of UPA in UC. Further research should be done regarding its effectiveness and safety in special populations, including pregnancy, breastfeeding, elderly population, or patients with acute severe UC. 20
Conclusions
UPA is a highly selective JAK 1 inhibitor, which is effective in managing moderate-to-severe UC. It is a welcome addition to the therapeutic armamentarium. There are no current predictive factors for UPA response; however, it can be considered an option for rapid symptomatic control in bio-naïve cases or in patients that have failed previous therapies. As an oral, once-a-day extended-release pill, it will be appreciated by patients due to convenience. Proper patient profiling is essential when considering UPA, and regular close monitoring is recommended to detect and avoid potential side effects.
