Vedolizumab use after failure of TNF-α antagonists in children and adolescents with inflammatory bowel disease

Inflammatory bowel disease (IBD), including Crohn’s disease (CD), ulcerative colitis (UC), and unclassified IBD (IBD-U), is a multifactorial disease that results from an intricate interplay between genetic predisposition, altered immune response, changes in the intestinal microbiota, and environmental factors. Together, these factors contribute to the destruction of the intestinal epithelial barrier, increased gut permeability, and an influx of immune cells [1, 2].

Treatment consists of induction, maintenance therapy and, during a flare, intensified therapy, with the aim of achieving and maintaining clinical remission. In paediatric CD, but not UC, exclusive enteral nutrition therapy is a first-line treatment and is as effective as corticosteroids, with less side effects [3]. Persistent unsatisfactory results from treatment with immuno-modulators (such as azathioprine or 6-mercaptopurine) or mesalazine and corticosteroids have led to the development of biologicals in the treatment of IBD. The most common biologicals, infliximab and adalimumab, work by inhibiting the proinflammatory cytokine, tumour necrosis factor alpha (TNF- α), and are effective in moderate to severe paediatric CD and UC [4, 5]. Unfortunately, up to a third of patients do not have a primary response, and primary responders may develop a loss of efficacy [6]. Patients who are switched to a second anti-TNF- α drug are less likely to show a good response and may develop antibodies [7].

Several novel agents are currently being investigated, including vedolizumab, golimumab, and other biologicals [8]. Vedolizumab is a humanized α4β7-integrin antagonist, characterized by a gut-selective mechanism of action. By binding to surface-expressed α4β7 integrin, it inhibits T-lymphocyte migration into inflamed intestinal tissue. Its efficacy and safety has been evaluated in adult patients with moderate to severe UC and CD [9‐12]. Initial data suggest that vedolizumab exerts its full effect after 14 weeks. Although vedolizumab was not more effective than placebo after prior TNF- α antagonist failure in adult patients with CD at week 6 (remission rate, 15.2%), a higher proportion of remission was achieved with vedolizumab than with placebo at week 10 [13]. Experience with vedolizumab in paediatric patients is limited, as its use is off-label, when available conventional therapy (TNF- α antagonists or immuno-modulators) has failed. Currently, only two case series from the United States and one from Europe have been published. These studies indicate a clinical benefit of vedolizumab in the induction and maintenance of remission in paediatric IBD, with a very good safety profile [14‐16].

The aim of this study is to complement the limited existing data by reporting the efficacy and safety of vedolizumab in a cohort of Austrian paediatric patients with IBD.

The median age of the 12 enrolled patients at baseline was 15 years (range, 8–17 years). All patients were previously treated with infliximab, 11 patients (92%) were previously treated as well with adalimumab, and one patient (8%) was previously treated with golimumab as a third biologic agent. The main reason for medication discontinuation was loss of response (58% for infliximab, 64% for adalimumab). Demographic data at start of vedolizumab treatment are shown in Table 1. Among the 7 patients with CD, one had neurofibromatosis type I as a second underlying disease, one showed extra-intestinal manifestations (arthritis), and one had a perirectal fistula. None of the patients underwent any surgery prior to vedolizumab treatment; however, one patient with UC underwent total colectomy after 60 weeks of treatment.

One patient discontinued therapy due to a severe general systemic allergic reaction at the second infusion of vedolizumab. This patient was included in the demographic data and short individual description in Tables 1 and 3, but was excluded in further investigations and outcome data (eg. Table 2). All of the remaining 11 patients received at least 5 vedolizumab infusions and completed 22 weeks of surveillance. Total observation and evaluation of the clinical course was maintained until week 38.

The present study adds to the scarce literature on vedolizumab use in paediatric IBD, and suggests that vedolizumab is safe and might provide a clinical benefit. Our cohort consisted of patients with refractory disease, in whom conventional therapy had failed. At week 14, clinical remission was achieved in 1/7 of the patients with CD and 4/5 of the patients with UC. Only one patient discontinued therapy due to a severe general systemic allergic reaction with dyspnoea at the second infusion of vedolizumab. These results are consistent with previous paediatric investigations that showed a significant reduction in activity scores only in patients with UC, with sample sizes of 64, 52, and 21, respectively [14‐16].

At week 14, 4/5 patients with UC achieved clinical remission, which is comparable to the rates reported in Singh et al. and Conrad et al. (week 14 clinical remission, 13/17 and 3/4 respectively), but is higher than that reported in Ledder et al. (week 14 clinical remission rate, 16/34). Although a significant reduction in the PUCAI scores was observed by week 2, 4/5 patients had concomitant corticosteroid therapy at this time. At the last follow-up, two patients in the UC group were still in remission; one patient had to undergo a total colectomy due to a severe disease course and loss of response, despite shortening of the dosing interval; and two patients presented a variable disease course. These results are consistent with the corticosteroid-free remission rate of 16/41 in paediatric UC after 12 months of therapy reported in Ledder et al. [16]. Additionally, the colectomy rate during vedolizumab treatment is similar to that in all previous paediatric reports (1/5, 4/22 and 6/41 respectively) [14‐16]. In previous reports, paediatric patients with CD had a poorer response to vedolizumab therapy than patients with UC [14‐16]. In the present study, the patients in the CD cohort had longer and more severe disease courses compared to those in the UC cohort before the introduction of vedolizumab. This may contribute to the lower response rate in the patients with CD. Shelton et al. suggested that the transmural healing process in CD might take longer compared to that in UC [21]. In addition, TNF-α antagonists downregulate MAdCAM-1 expression; thus, patients with CD with prior anti-TNF exposure may require more time to respond to treatment [13, 22].

The RTE values in the CD cohort remained similar across the observation period, and were lower in weeks 14 (0.46) and 38 (0.44) compared to that at baseline (0.56). Even more pronounced, RTEs in the UC cohort dropped markedly from 0.75 at baseline to 0.25 in weeks 14 and 38. In conclusion, the RTE indicates improvements in disease activity scores in the UC cohort, but not in the CD cohort.

Regarding secondary outcomes, growth failure and nutritional status in IBD patients, especially in children and adolescents, is a big concern [23]. Particularly as body image concerns impact the quality of life in this cohort [24]. In our data substantial weight gain in UC and a trend in CD was observed, while in the largest paediatric cohort study on vedolizumab thus far, no improvement in height velocity and no significant weight gain were observed during treatment [16]. Height and weight were not presented in two other paediatric studies [14, 15].

An improvement in laboratory markers, such as albumin and haemoglobin, were observed in both groups, relative to baseline values. In previous paediatric studies, no relevant treatment-related changes were found in HKT, albumin, and CRP levels [14]. Similarly, Singh et al. reported a treatment-related decrease in CRP levels in patients with CD, while CRP values were normal at baseline in the patients with UC, and albumin and HKT levels were normal at baseline and did not change over time [15].

In the largest paediatric cohort study on vedolizumab thus far, no improvement in height velocity and no significant weight gain were observed during treatment, but a decrease in faecal calprotectin was observed at the last follow-up compared to that at baseline [16].

Regarding safety concerns and severe adverse events, one patient, who had previously shown a systemic reaction to infliximab and a local skin reaction to adalimumab, had a severe systemic allergic reaction at the second infusion of vedolizumab. No other infusion reactions or severe adverse events were observed. Consistent with these results, minor adverse events, but no severe adverse events or allergic reactions, were reported in the existing paediatric studies. Similarly, GEMINI I and II reported no severe adverse events in adult patients, while GEMINI III reported one (< 1%) drug-related severe adverse event [11, 13, 25].

The present study has some limitations. The study was retrospective in nature and the sample size was small, even though we included all paediatric patients who received vedolizumab in Austria from November 2014 to August 2016; thus, firm conclusions on the effect of vedolizumab cannot be drawn. In particular, it should be emphasized that the p values and the corresponding statements about statistical significance included in the present manuscript must not be over-interpreted, since the sample sizes are very limited. However, 8 of our patients were followed for at least 52 weeks, which is comparable to the study by Ledder et al. [16]. Due to the retrospective character of the study, it was not possible to include faecal calprotectin in the CD cohort and ESR values. Prospective long-term studies, like the currently enrolling VEDOKIDS trial are required to define the full safety profile and adequate dosing for vedolizumab treatment in children with IBD [22].

Our results suggest that vedolizumab may induce a clinical response and an improvement in soft markers, like weight gain. However, in children with a severe disease course or steroid-dependency, a loss of response to vedolizumab occurred quite often, and vedolizumab use was not successful in preventing a colectomy in one child with UC.

In conclusion, the present study demonstrates that vedolizumab might be an effective treatment for individual paediatric patients with IBD who are unresponsive, intolerant, or experience a loss of efficacy in other therapies, including TNF-α antagonists. However, vedolizumab appears to be more effective in paediatric UC than in paediatric CD.