Over the past decade, worldwide incidence of inflammatory bowel disease (IBD) has been increasing [1], including in the pediatric population. While IBD primarily affects young adults, 15% to 25% of cases occur in childhood [2,3]. By 5 years post-diagnosis, 18% of pediatric IBD patients have been found to require surgical intervention [4]. Despite the well-established efficacy of anti-tumor necrosis factor (TNF) therapies such as infliximab and adalimumab, some patients need a change in therapy due to allergic reactions, adverse events, as well as primary and secondary loss of response, highlighting the necessity for therapies acting by different mechanisms [5].

Ustekinumab, a monoclonal antibody targeting the p40 subunit of interleukin-12 and interleukin-23, has been approved to treat adults with active psoriatic arthritis or moderately to severely active Crohn’s disease (CD), as well as to treat both adults and children aged 12 or older with moderate to severe plaque psoriasis [6].

Thus far, in the adult CD population, ustekinumab has demonstrated variable clinical response rates ranging from under 40% to as high as 84% [7,8]. When successful, it has been shown to induce both clinical remission and mucosal healing, including in patients with refractory CD who have already received an anti-TNF [9-15]. At present, limited data exists on ustekinumab use in pediatric CD, showing variable results [16-20].

We retrospectively describe 11 adolescent patients who received ustekinumab at the Montreal Children’s Hospital with the goal of assessing its efficacy in inducing clinical, biochemical, and endoscopic remission.

A retrospective chart review was performed for all patients at the Montreal Children’s Hospital with CD who received ustekinumab between December 2015 and July 2018. Baseline data, information on prior treatment and response, indications for starting ustekinumab, clinical information, and laboratory parameters pre- and post-therapy were abstracted. Duration of disease, location, and phenotype was recorded according to the Montreal classification [21]. Response to therapy was assessed using laboratory markers, clinical data, and both endoscopy and pathology reports when available. Clinical response was defined as a decrease in score on the abbreviated Pediatric Crohn’s Disease Activity Index (abbrPCDAI), a version of the PCDAI used in retrospective studies when all data required to calculate the full PCDAI score is not available. Specifically, the variables removed were: height, hematocrit, albumin, and erythrocyte sedimentation rate (ESR). [22]. ESR could not be utilized, as it is not part of standard evaluation at our center. The abbrPCDAI has been found to demonstrate sufficient correlation with the full PCDAI [23]. Scores range from 0 (clinical remission) to 70 (severe, active disease), with the suggested cutoff points being <10 (remission), 10 to 15 (mild disease), 16 to 25 (moderate disease), and >25 (severe disease) [23]. Patients who attained abbrPCDAI scores in the remission range (<10) at any point throughout the course of their treatment were classified as responders, with the exception of 2 patients who had low abbrPCDAI scores but severe disease on endoscopy or magnetic resonance enterography (MRE) both before and after ustekinumab therapy. These 2 patients, alongside those with abbrPCDAI scores ≥10, were classified as nonresponders.

Two patients were excluded from the data analysis. One had less than 6 months of follow-up data, and in another, there was a discrepancy between the biochemical markers (normal) and clinical symptoms, and thus it was not possible to accurately determine if the patient had active disease since endoscopy/imaging was not performed.

Certain patients’ medical charts were reviewed independently for another study (Chavannes et al.) [18]. In the cases of these patients, the overlap predominantly consists of baseline data, with the current study including additional outcome variables and longer follow-up. This study was approved by the Research Ethics Board of the McGill University Health Centre (approval No. 2019-4728). This study is a retrospective study using medical record review and so informed consent was waived.

Two of 11 patients who had previously failed anti-TNF therapy were able to attain states of clinical remission within 2 to 7 months on ustekinumab, maintaining remission across 6 and 22 months of follow-up data respectively. Of the remaining 9 patients, 6 were primary nonresponders (defined as a lack of clinical improvement following induction therapy [24]) and 3 showed initial response with a subsequent secondary loss of response.

Weight gain was seen in all but 1 patient, including those not labeled as having responded to ustekinumab. Notably, a confounding factor may be found in the fact that 2 out of 6 (33.3%) of these nonresponders were taking oral corticosteroids at the time. Only 1 patient experienced significant clinical worsening while taking ustekinumab; patients generally either improved or maintained a consistent symptom burden, which may have allowed for the weight gain observed.

Assessing normalization of inflammatory markers can be a challenge, as it has been found that 20% to 25% of patients with CD do not exhibit increased CRP even when their disease is active due to a genetic polymorphism affecting CRP production [25]. In our study, 5 patients did not have an elevated CRP prior to initiating ustekinumab treatment despite having elevated FCP (ranging from 384 μg/g to >2,100 μg/g). These patients have never had a documented elevated CRP. Of the clinical responders, 3 out of 5 had elevated CRP values prior to initiating ustekinumab therapy, all of which normalized within 1 month of clinical improvement. Even 1 patient who did not benefit from ustekinumab clinically experienced a rapid decrease in CRP within 2 months, thus demonstrating that ustekinumab is capable of improving inflammatory markers.

In the treatment of CD, achieving mucosal healing is the ideal goal, as this has been associated with better outcomes, including a decrease in both corticosteroid use and hospitalizations. The International Organization for the Study of Inflammatory Bowel Diseases (IOIBD) defines mucosal healing as the absence of visible ulcerations and erosions on endoscopy [26]. Sufficient endoscopic data was only available for 4 patients, 2 of whom responded to ustekinumab. Based on the IOIBD definition, ustekinumab did demonstrate the ability to induce mucosal healing in 1 of the patients who responded to therapy, also resulting in clear endoscopic improvement in both a responder and nonresponder. Ustekinumab has previously been found to cause mucosal healing in adults [11], but this has not yet been documented in the pediatric population.

Concomitant use of immunomodulators was previously found to be a potential predictive variable for clinical benefit of ustekinumab at 3 months [11], however we did not find any such relationship. Four of our patients were taking either methotrexate or azathioprine at the time of induction—one responded to ustekinumab and the other 3 did not. However, given the limited sample size a definite conclusion cannot be drawn regarding combination therapy.

Previous studies examining ustekinumab use in the pediatric population are few in number. Two case studies, documenting the clinical progress of 1 patient each with severe and refractory CD, showed variable results, with ustekinumab inducing remission in 1 patient and failing in the other [17,20]. A chart review examined the potential for clinical improvement on ustekinumab in 4 pediatric patients, 2 of whom experienced clinical worsening and 2 of whom demonstrated clinical improvement [16]. Most recently, a multi-center cohort study examined 44 pediatric CD patients retrospectively, finding that at 12 months, 38.6% achieved clinical remission and 47.8% achieved clinical response [18]. Another recent study found that amongst 52 children with either CD or ulcerative colitis initiating ustekinumab, 50% of those previously exposed to anti-TNF treatment were in steroid-free remission by week 52 [19].

Our results demonstrate the potential for mucosal healing in pediatric patients receiving ustekinumab. Furthermore, while in the other pediatric studies, all patients received maintenance doses of ustekinumab every 7–8 weeks, doses in our study were shown to be given as frequently as every 4 weeks in an attempt to optimize therapy prior to labeling a patient as a nonresponder, with no adverse effects noted. Among the 8 patients who had their treatment interval shortened, 3 patients (37.5%) showed a clinical response.

This data suggests that ustekinumab has the potential to result in clinical and endoscopic improvement, as well as decreased inflammatory markers. However, only 1 patient in this study was proven to maintain remission on ustekinumab for longer than 1 year, with this patient still clinically stable after 22 months of therapy. One responder only had 6 months of follow-up data available, while the others experienced secondary losses of response within 8–12 months of induction. Previous results have suggested with an even smaller sample size that the benefit of ustekinumab has not lasted longer than 15 months [20]. Ideally, the next step will be to conduct a large prospective randomized control trial. Should this finding persist with a larger sample size, the next question will be whether anything can be done to optimize treatment and allow it to have a sustained effect.

This study was limited both by its small sample size, which decreased generalizability, and by the fact that it was retrospective in nature. Being retrospective, we were not able to control for certain differences amongst patients and were not able to collect all desired data at specific and regular time intervals. As a result, we did not have FCP and endoscopic data both pre- and post-ustekinumab for all patients, nor did we have the regular laboratory assessments needed to calculate the full PCDAI score. Furthermore, as described in Table 3, patients underwent several different induction protocols, which may have confounded the findings.

Future research also ought to investigate whether or not ustekinumab has a greater treatment effect in pediatric patients who are anti-TNF-naïve when compared to patients who have already failed TNF antagonist therapy, as has been found to be the case with vedolizumab [27]. Data from the UNITI-2 study, with 69% of patients who were completely TNF antagonist-naïve, showed that these patients had higher rates of absolute response and remission than those who had previously failed anti-TNF therapies [28]. Similarly, in our population that was not biologic-naïve, the response rate was much lower. However, by shortening the interval, some patients were able to achieve a clinical response.

Ustekinumab is a promising treatment option not only for adults with CD, but for children as well. Future research is needed to determine factors that influence response to therapy and which patients are most likely to benefit from this medication in order to optimize treatment strategies.