Efficacy of Dose Escalation of Ustekinumab in Inflammatory Bowel Disease: A Systematic Review and Meta-Analysis

Inflammatory bowel disease (IBD) is a chronic, non-infectious, relapsing-remitting inflammatory condition of the gastrointestinal tract, and it two conditions, namely, Crohn’s disease (CD) and ulcerative colitis (UC). CD is a transmural inflammation of the gastrointestinal tract characterized by deep, patchy skip lesions [1]. Patients can experience severe complications including the formation of strictures and fistulas between the bowel and other adjacent organs. Fistula formation is mostly associated with disease and treatment burden, which negatively impacts patients’ quality of life [2]. UC, on the other hand, the inflammation is limited to the colonic mucosa alone. IBD is a major cause of morbidity and mortality, with nearly 5 million global cases being reported annually [3, 4].

Over the last 2 decades, many biologics have become available for the treatment of IBD. These biologics are employed for the management of IBD cases that do not respond to conventional pharmacological agents. The biological options for IBD encompass anti-TNF inhibitors, anti-integrin inhibitors (vedolizumab), and anti-interleukin-12/23 inhibitors (ustekinumab) [5, 6].

Ustekinumab, a fully humanized IGk monoclonal antibody targeting the P40 subunit of interleukin-12/23 was approved by FDA in 2016 for the treatment of CD [6‒8]. Studies show that genes encoding this interleukin, along with findings of genome-wide association, have been associated with the pathogenesis of CD [7, 9]. Ustekinumab blocks this inflammation pathway by inhibiting the receptors of IL-12/23 on the immune cells and thus shows efficacy for moderate-to-severe CD and UC [3, 10]. The pharmacokinetic and pharmacodynamics are significantly affected by various biologic factors which change clinical response to the treatment modality [11]. The patients receiving such biologic therapy are categorized on the basis of response rate. Those patients who fail to respond initially are classified as primary non-responder. Conversely patients who respond well to biologic therapy initially but over time show a decrease in the therapeutic effect of biologics are classified as secondary non-responder. This is called loss of response, which is a clinical challenge for gastroenterologists. In such patients, different treatment strategies are employed, such as switching to another drug, surgery, dose escalation via a decrease in dosage interval, increasing the dosage, or intravenous (IV) reinduction [11]. With the recent expiry of the ustekinumab patent, development of several biosimilars is expected to make ustekinumab a cost-effective option for IBD. This development promises to make dose escalation more appealing to combat partial or complete loss of response. This systematic review and meta-analysis pools the real-world data reporting the efficacy of dose escalation of ustekinumab in patients with IBD.

This systematic review and meta-analysis adheres to the guidelines established by the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) and was registered with PROSPERO (CRD42024550966) [12].

A systematic literature search was conducted on PubMed, Embase, Clinicaltrails.gov, and Cochrane from inception to June 1, 2024, using Medical Subject Headings (MeSH) and keywords for “Inflammatory Bowel Disease” and “Ustekinumab.” We also hand-searched conference proceedings of the American College of Gastroenterology for the last 5 years. Bibliographies of all the included articles were also searched for relevant articles. The detailed search strategy for one database is presented in online supplementary Table 1 (for all online suppl. material, see https://doi.org/10.1159/000543831).

We imported all the references retrieved by our search into Endnote v20 and the duplicates were removed. Two authors (A.A. and A.H.B.G.) independently screened the titles and abstracts of all the studies and removed those that did not fulfil our inclusion criteria. The remaining articles were evaluated against the eligibility criteria by reviewing their full texts. Disagreements and conflicts were discussed with a third author (M.E.U.R.) and resolved. The eligibility criteria were: (1) Population: patients with IBD; (2) Intervention: dose escalation of ustekinumab from a maintenance regimen of 90–95 mg subcutaneously every 4–12 weeks; (3) Study Designs: cohorts.

Data were extracted independently by two authors (A.Q.K. and M.S.) into a pre-defined data collection form on Microsoft Excel. In case of any discrepancies, a third author (A.T.) was consulted. The data extracted from the included studies comprised of study characteristics (first author, publication year, study setting, study design, and sample size), patient characteristics (age, sex, duration of IBD, and follow-up duration), and treatment characteristics (maintenance dose of ustekinumab, intensified dose of ustekinumab, and time to dose escalation). Intensification strategies were categorized as either interval shortening (increase in frequency of administration) or IV reinduction (switching from subcutaneous to IV ustekinumab). Outcomes assessed were clinical response (meaningful improvement in symptoms), clinical remission (absence or near absence of symptoms), steroid-free remission (absence or near absence of symptoms without use of steroid therapy), endoscopic response (improvement in endoscopic findings), endoscopic remission (resolution or near resolution of endoscopic findings of active disease), normalization of CRP, normalization of faecal calprotectin, remission of perianal disease, and progression to surgery following ustekinumab dose escalation. Clinical response and clinical remission were the primary outcomes.

The Newcastle-Ottawa Scale (NOS) was used for the quality assessment of each of the included studies [13]. Overall bias was calculated based on the three domains of the NOS. The NOS has three categories; selection (4 items), comparability (1 item), and outcome (3 items). The included studies were single-arm cohorts so we excluded 2 items; namely, comparability and selection of non-exposed cohort. After this modification, each study could score between 0 and 6. The risk of bias was labelled as high, unclear, or low. Two authors (F.S. and S.Z.) assessed the risk of bias independently. A third author (A.H.) was consulted in case of any discrepancies.

Meta-analyses of proportions were performed for all outcomes after variances were stabilized using the Freeman Tukey double arcsine transformation. The Der Simonian method with a random-effects model was used to pool the prevalence values for categorical outcomes, and the pooled results were visually depicted using forest plots. The I² statistic was used to assess the heterogeneity between the included studies. Cochrane Handbook for Systematic Reviews of Interventions, section 10.10 was used to interpret the I² values [14]. Subgroup analyses were conducted based on intensification strategy (interval shortening versus IV reinduction). Statistical analyses were performed using MetaXL version 5.3 [15].

The literature search retrieved 3,207 studies. After the exclusion of duplicates, reviews, and ineligible articles, we included a total of 28 studies in our review [6, 16‒42]. The literature screening process is summarized in Figure 1.

We reviewed 28 studies, of which seven were conducted in the USA, four in Spain, three in the Netherlands, Canada, and China, two in the United Kingdom, and one each in Singapore, France, Switzerland, Scotland, Slovenia, and Israel. The publication years ranged from 2018 to 2024. A total of 2030 patients (1,858 with CD and 172 with UC) were included. The median age ranged from 21.5 to 51.5 years. The characteristics of included studies are summarized in Table 1.

The quality assessment of included studies is presented in online supplementary Table 2. Due to a lack of non-exposed cohort in our study, the compatibility domain was not applicable. Out of a total of 6 points, seven studies were awarded 6, sixteen studies were awarded 5, four studies were awarded 4, and 1 study was awarded 3 points.

Eighteen studies reported data on clinical response. The analysis included 1,218 patients, of whom 692 achieved a clinical response, resulting in a pooled prevalence of 55% (95% CI: 46–65%, I² = 91%) (Fig. 2). Fourteen studies reported clinical response for CD patients alone. Out of 1,067 patients, 639 achieved a clinical response, yielding a pooled prevalence of 59% (95% CI: 49–68%, I² = 90%) (online suppl. Fig. 1).

In a subgroup analysis on the basis of intensification strategy, the pooled prevalence for clinical response for interval shortening was 55% (95% CI: 43–67%, I² = 89%). For IV reinduction, the pooled prevalence of clinical response was 52% (95% CI: 28–76%, I² = 95%) (online suppl. Fig. 2).

Fifteen studies (1,041 patients) were included in the analysis for clinical remission. Out of these, 524 achieved clinical remission, resulting in a pooled prevalence of 51% (95% CI: 42–59%, I² = 87%) (Fig. 3). The prevalence of CD patients who achieved clinical remission was 52% (95% CI: 42–61%, I² = 86%) (online suppl. Fig. 3).

The subgroup analysis of clinical remission by intensification strategy revealed a pooled prevalence of 50% (95% CI: 36–64%, I² = 91%) for interval shortening. For IV reinduction, the pooled prevalence was 54% (95% CI: 43–64%, I² = 60%) (online suppl. Fig. 4).

The analysis showed that 244 out of 484 IBD patients in 10 studies achieved endoscopic response, with a pooled prevalence of 54% (95% CI: 42–65%, I² = 82%) following UST dose intensification (online suppl. Fig. 5). For CD patients, endoscopic response had a pooled prevalence of 58% (95% CI: 48–68%, I² = 57%) (online suppl. Fig. 6). The subgroup analysis by intensification strategy revealed a pooled prevalence of 51% (95% CI: 33–69%, I² = 87%) for interval shortening and a pooled prevalence of 62% (95% CI: 44–78%, I² = 59) for IV reinduction (online suppl. Fig. 7).

Eleven studies reported this outcome. 152 out of 444 patients achieved endoscopic remission and the pooled prevalence was 33% (95% CI: 24–42%, I² = 73%) (online suppl. Fig. 8). Endoscopic remission was attained by 99 out of 319 CD patients, yielding a pooled prevalence of 30% (95% CI: 21–40%, I² = 69%) (online suppl. Fig. 9). The pooled prevalence for interval shortening was 30% (95% CI: 21–41%, I² = 67%) while the pooled prevalence for IV reinduction was 47% (95% CI: 23–71%, I² = 62%) (online suppl. Fig. 10).

Fourteen studies involving 779 patients with IBD reported steroid-free remission following UST dose intensification. Among these patients, 428 achieved this outcome, resulting in a pooled prevalence of 54% (95% CI: 45–64%, I² = 84%) (online suppl. Fig. 11). For CD patients, the pooled prevalence was 53% (95% CI: 40–65%, I² = 87%) (online suppl. Fig. 12).

In the subgroup analysis by intensification strategy, the pooled prevalence was 49% (95% CI: 39–60%, I² = 83%) for interval shortening and 75% (95% CI: 55–92%, I² = 63%) for IV reinduction (online suppl. Fig. 13).

Data from 11 studies were analysed to assess the normalization of CRP levels following UST dose intensification; 302 out of 702 patients achieved this outcome, resulting in a pooled prevalence of 42% (95% CI: 33–52%, I² = 84%) (online suppl. Fig. 14). The pooled prevalence of CRP normalization was 40% (95% CI: 29–52%, I² = 87%) for CD patients after UST dose intensification (online suppl. Fig. 15).

In the subgroup analysis by intensification strategy, the pooled prevalence was 37% (95% CI: 27–48%, I² = 79%) for interval shortening and 57% (95% CI: 47–68%, I² = 28%) for IV reinduction (online suppl. Fig. 16).

Eight studies (651 patients) reported this outcome; 55 patients required surgical intervention, resulting in a pooled prevalence of 8% (95% CI: 5–12%, I² = 57%) (online suppl. Fig. 17). The pooled prevalence for CD patients was 8% (95% CI: 4–13%, I² = 63%) (online suppl. Fig. 18). The subgroup analysis by intensification strategy revealed a pooled prevalence of 10% (95% CI: 6–14%, I² = 54%) for interval shortening and 3% (95% CI: 1–8%, I² = 0%) for IV reinduction (online suppl. Fig. 19).

After UST dose intensification, the pooled prevalence of perianal disease remission in CD patients was 32% (95% CI: 22–43%, I² = 39%) (online suppl. Fig. 20). In the subgroup analysis by intensification strategy, the pooled prevalence was 34% (95% CI: 22–46%, I² = 43%) for interval shortening and 26% (95% CI: 0–67%, I² = 53%) for IV reinduction (online suppl. Fig. 21).

Eight studies reported on faecal calprotectin levels in IBD patients after UST dose intensification. Out of 278 patients, 155 achieved this outcome, resulting in a pooled prevalence of 54% (95% CI: 41–67%, I² = 75%) (online suppl. Fig. 22). For CD patients, the pooled prevalence was 54% (95% CI: 39–68%, I² = 74%) (online suppl. Fig. 23). For interval shortening, the pooled prevalence was 61% (95% CI: 44–77%, I² = 57%) and for IV reinduction, it was 55% (95% CI: 34–75%, I² = 73%) (online suppl. Fig. 24).

This systematic review and meta-analysis pooled 28 cohort studies to report the efficacy of dose escalation of ustekinumab in IBD patients. In our study, more than half of the patients achieved clinical remission, and one-third achieved endoscopic remission. Normalization of CRP and faecal calprotectin levels were achieved by 42% and 54% of the subjects, respectively. Only 8% of patients required IBD-related surgery following dose escalation.

In recent times, the dose escalation of ustekinumab has emerged as a crucial strategy in managing IBD, particularly when patients exhibit persistent disease activity. This approach is typically initiated in response to key indicators such as continuous endoscopic inflammation, elevated biomarkers like C-reactive protein (CRP) and faecal calprotectin, and sub-therapeutic serum drug levels [10]. The dose escalation strategies for ustekinumab include IV reinduction or interval shortening from every 12th week to every 8th, 6th, or 4th week [43]. Overall, interval shortening was the most frequent strategy used for dose escalation in our meta-analysis.

The primary outcomes included in our review were clinical response and clinical remission. Clinical response was reported in 55% (95% CI: 46–65%) of patients after dose escalation. A meta-analysis by Meserve et al. [11] found that 55% of patients achieved a clinical response following dose escalation, similar to our findings. The subgroup analysis based on the dose intensification strategy reported that for patients undergoing interval shortening, the clinical response was 49% (95% CI: 38–61%). In contrast, IV reinduction had a higher clinical response of 54% (95% CI: 43–64%). Notably, Tang et al. [39] specifically examined the efficacy of IV reinduction, reporting a remarkably high clinical response rate of 75.5%. This highlights the potential effectiveness of the IV reinduction approach.

The pooled prevalence of clinical remission was 51% (95% CI: 42–59%). Yang et al. [5] reported that 58% of patients reached clinical remission following dose escalation, suggesting a similar trend in treatment efficacy. For patients with CD, the pooled prevalence of clinical remission was 52% (95% CI: 42–61). This finding underscores that CD patients respond similarly to the broader IBD population in achieving remission after dose escalation. Further analysis showed that interval shortening led to a 50% clinical remission rate (95% CI: 36–64%). Ollech et al. [33] reported that 50.9% of patients achieved clinical remission with this approach, aligning closely with our pooled analysis. Conversely, for IV reinduction the pooled prevalence was 54% (95% CI: 43–64%). This finding was in line with Huinink et al. [28] who reported that 45% of patients achieved clinical remission following IV reinduction. Interestingly, Kopylov et al. [31] reported comparable remission rates between the two strategies.

Analysis of secondary outcomes revealed that 33% (95% CI: 24–42%) of patients achieved endoscopic remission. In a subgroup analysis, we found that in the IV reinduction group, 47% (95% CI: 23–71%) while in the interval-shortening group, 30% (95% CI: 21–41%) achieved endoscopic remission. This suggests that the IV reinduction method has more potency, despite the considerably wide confidence intervals. Yao et al. [40] noted that ustekinumab trough concentrations were higher in the IV reinduction group compared to the interval-shortening group, potentially justifying a superior endoscopic remission rate with IV reinduction. Additionally, Hanžel et al. [26] demonstrated that an ustekinumab serum concentration of 6.00 mg/L at endoscopy strongly predicted endoscopic remission, with a sensitivity and specificity of 75%, and 82%, respectively. This suggests that measuring drug concentrations following dose escalation may serve to predict endoscopic response and can guide clinicians in adjusting treatment plans.

Our analysis reported that 54% of IBD patients (95% CI: 45–64%) were weaned off from steroids after dose escalation. With examining intensification strategies, we noted that 49% (95% CI: 39–60), of patients who underwent interval shortening achieved steroid-free remission. This contrasts with IV reinduction, which had a higher success rate, with 75% (95% CI: 55–92) of patients reaching steroid-free remission. Meserve et al. [11] reported only 40% (95% CI: 21–61%) of patients achieving steroid-free remission after interval shortening. Steroid-free remission is dependent on multiple patient characteristics. Olmedo Martín et al. [34] identified several factors that play a key role in achieving steroid-free remission after dose escalation. Multivariate analysis pointed out that perianal involvement and ongoing steroid use at the initiation of dose escalation were significant predictors of failure to achieve steroid-free remission.

CRP normalization was observed in 42% of the patients (95% CI: 33–52%). Detailed subgroup analysis indicated that CRP normalization was achieved in 57% of patients (95% CI: 47–68%) in the IV reinduction group, while only 37% (95% CI: 27–48%) of patients in the interval shortening group experienced CRP normalization. Furthermore, a cohort study by Yao et al. [40] found that CRP normalization rates were 50.5% in the interval-shortening group and 56.5% in the IV reinduction group. This study also highlighted a significant correlation between CRP normalization and ustekinumab trough concentration, noting that patients with ustekinumab trough concentrations above 4.8 µg/mL had an 85% prevalence of CRP normalization.

Several studies have indicated that dose escalation is frequently employed in patients with CD who have perianal disease. Studies have found a significant association between the presence of perianal disease and the likelihood of dose escalation (HR 3.00; 1.05–8.59, p = 0.04) [21, 32]. This suggests that the presence of perianal disease often necessitates increased dosing to manage symptoms effectively. In our study, we found that 32% (95% CI: 22–43) of CD patients achieved remission. Similarly, a study by Ollech et al. [33] reported that 45% of patients experienced remission of perianal disease after dose escalation.

Our findings should be interpreted with caution due to several limitations. Firstly, only cohort studies were included, which may compromise the robustness of the findings. Secondly, the majority of included studies had a retrospective study design. These factors contribute to a high risk of bias. The presence of high heterogeneity among the included studies poses a significant challenge. The subgroup analysis did not adequately explain the cause of high heterogeneity. Thirdly, patients with both CD and UC were included, which could contribute to heterogeneity. However, separate analysis was undertaken for CD to minimize heterogeneity due to type of IBD. Fourthly, clinical, endoscopic and biochemical response were judged by the included studies using several different criteria, which could significantly contribute to the heterogeneity observed in our analyses.

Nevertheless, this study has several merits. The inclusion of cohort studies provided real-world evidence, reflecting everyday clinical practices and patient populations including a wide range of participants, capturing diverse demographics, and treatment responses.

In conclusion, this systematic review and proportional meta-analysis highlight that dose escalation of ustekinumab is an effective way to manage IBD in patients who do not respond to standard doses. A significant proportion of these patients achieve clinical and biological remission with escalated doses. IV reinduction may potentially be associated with better endoscopic outcomes, remission from steroid and perianal disease, and normalization of CRP. However, high-quality, large-scale trials are needed to evaluate the long-term safety and efficacy of dose escalation of ustekinumab for use in IBD management and to clarify if IV reinduction is the most optimal intensification strategy.

Open access funding provided by the Qatar National Library.

This study was exempt from requiring ethical approval on account of being a systematic review. This study was exempt from requiring written informed consent on account of being a systematic review.

The authors have no conflicts of interest to disclose.

Open access funding was provided by the Qatar National Library. The funder had no role in the design, data collection, data analysis, and reporting of this study.

Mohammad Ebad Ur Rehman: conceptualization, formal analysis, writing – original draft, and writing – review and editing. Ammara Tahir: conceptualization, methodology, writing – original draft, and writing – review and editing. Amna Hussain: conceptualization, methodology, and writing – original draft. Aizaz Ali and Abu Huraira Bin Gulzar: methodology, formal analysis, and writing – original draft. Abdul Qadeer Khan, Maha Sajjad, Shahroon Zahid, Fatima Shahid, and Ummara Aslam: data curation, methodology, and writing – original draft. Talha Bin Yasin: data curation, methodology, writing – review and editing. Aqsa Bilal and Tehreem Fatima: data curation, formal analysis, and writing – original draft. Muhammad Sheraz Hameed: conceptualization and writing – review and editing. Tehseen Haider, Abdulqadir J. Nashwan, and Sajeel Saeed: formal analysis and writing – review and editing.

All data used in this study will be provided by the corresponding author on reasonable request.