Introduction
Inflammatory bowel diseases (IBDs), including both ulcerative colitis (UC) and Crohn’s disease (CD), are considered relapsing and remitting immune-mediated chronic inflammatory gastrointestinal diseases. [1, 2]. Over many years, patients suffering from these diseases were treated using classic medical treatment, including anti-inflammatory drugs such as aminosalicylates and corticosteroids or immunomodulators such as thiopurines or methotrexate, mostly in a step-up protocol [3].
It was a great achievement when anti-tumor necrosis factor (anti-TNF) drugs were introduced into the treatment protocols for IBD. The need to use steroids has been decreased to a great extent. In addition, using anti-TNF agents stimulates the healing of intestinal mucosa, and reduces the frequency of patients’ hospitalization and surgery. All of these factors make the quality of life of IBD patients better than before [4].
Unfortunately, about 30% of IBD patients do not show a good response to induction with anti-TNF agents. Those patients are known as primary non-responders [5]. Moreover, in some settings, some patients lose their good response to anti-TNF drugs although they had a satisfactory response at the beginning of therapy. Those patients are described as secondary non-responders. Some other patients may develop a significant adverse drug effect. Both groups of patients discontinue their anti-TNF therapy [6].
More than one-third of patients who respond to anti-TNF at the beginning of therapy stop showing good responses after a while. This is explained by multiple factors, mainly the subtherapeutic plasma level of the drug. Also, antidrug antibodies may develop against drug molecules. Another explanation is that another cytokine could play a more crucial role in the pathogenetic mechanism of the disease [7].
Therefore, to regain remission in those secondary non-responder patients, we have to shift to another anti-TNF agent [8]. Some studies have assessed how effective a second anti-TNF agent is in UC or CD patients who did not experience disease remission when they received an anti-TNF drug for the first time [9].
Researchers have concluded in previous studies that the subsequent use of another anti-TNF drug is usually associated with an increased probability of primary treatment failure [10].
Although anti-TNF drugs were, for many years, the only biological drugs available to treat UC and CD patients, recently novel biological therapies have been approved for therapy. These novel drugs were designed to target different inflammatory pathways other than anti-TNF [11, 12].
Currently, these novel biological therapies that are used to treat IBD patients are vedolizumab, which is an anti-α4β7 integrin monoclonal antibody, and ustekinumab, which works as an anti-IL12/23 p40 antibody [12].
Using these novel agents as first- or second-line biological therapies for IBD patients, they have proven to have a good effect in the induction and maintenance of remission of the disease [13].
After understanding and studying the biology of T-lymphocytes and their active pathological role in mediating gut inflammation in IBD, some biological agents that act against T lymphocytes aggregating in the intestinal mucosa have recently been developed [14].
The new therapeutic target was to suppress leukocyte migration to the gut mucosa. Vedolizumab has achieved this target; it is the premium ‘gut-selective’ biological drug and the only one currently licensed [15].
Ustekinumab is an IgG1 monoclonal antibody that attaches to the P40 subunit expressed by the pro-inflammatory interleukins 12 and 23 [16].
It will be crucial in the future to understand the mechanisms behind the response of IBD patients to anti-TNF therapy. We aim to select and choose the most appropriate biologic agent for each patient [17].
Multiple randomized placebo-controlled trials have been carried out recently, to evaluate the effectiveness of vedolizumab and ustekinumab in IBD therapy. In these Phase 3 registration trials, nearly 50% of patients had not received any anti-TNF before [13]. In another setting, multiple observational cohort studies have confirmed that the majority of IBD patients treated using vedolizumab and ustekinumab were previously treated with anti-TNF drugs [18].
Nowadays, choosing the best second-line biological drug to treat patients who are non-responsive to anti- TNF agents is very challenging [19].
Aim
The current study aimed to evaluate the effectiveness of second-line biologics (ustekinumab and vedolizumab) after failure of an anti-TNF agent in inducing and maintaining remission over 12 months.
Material and methods
Study design
The study was observational and cross-sectional. Our study included adult patients between 18 and 60 years of age with confirmed diagnoses of UC and CD, who were treated using either ustekinumab or vedolizumab after stopping anti-TNF and manifested with an exacerbating disease at the time of starting ustekinumab or vedolizumab therapy. This study was carried out in our IBD outpatient clinic, Tropical Medicine Department, Ain Shams University Hospitals, in the period between January 2023 and January 2024.
Patients who did not agree to participate in the study and patients with any contraindication to therapy such as drug hypersensitivity, active infection, or active tuberculosis, were excluded from this study.
Primary non-response, secondary loss of response, or even intolerance to the given drug is considered failure of the first anti-TNF drug. From the start of the second-line medication until the end of the follow-up period (the last clinical visit) or the point at which therapy was stopped for any reason, we were monitoring the patients.
Study procedures
Baseline information, including gender, age, body mass index, smoking status, duration of the disease, and surgical history was gathered from the patients. Diarrhea, abdominal pain, fever, weight loss, rectal bleeding, constipation, tenesmus, and extra-intestinal signs including arthralgia, red eyes, backaches, were all clinical symptoms that were recorded. Additionally, previous usage of mesalamines, corticosteroids, and immunomodulators (thiopurines) for IBD was recorded.
Basic blood tests such as complete blood count (CBC), C-reactive protein (CRP), and erythrocyte sedimentation rate were performed for all patients. Also, intestinal infections were excluded by performing stool analysis and culture. All patients underwent a complete colonoscopy with numerous mucosal biopsies to accurately diagnose their conditions and determine the severity and extent of the disease.
Before starting medical treatment, patients were tested for viral infection and tuberculosis (HBsAg, HCV Ab, HB core IgG, and QuantiFERON TB gold). Type of initial anti-TNF agent (adalimumab or infliximab), first and last injection timing, reasons for termination, and combination with other immunosuppressants were noted.
Information on second-line therapy, including medication type (ustekinumab or vedolizumab), first and last injection timing, and whether corticosteroids or immunosuppressants were combined were documented.
Drug regimens
Patients received treatment plans that followed the approved drug’s guidelines. We began vedolizumab therapy with 300 mg intravenous infusions at weeks 0, 2, and 6 as induction therapy, and then every 8 weeks for maintenance therapy. Ustekinumab was administered intravenously as an induction dose of 6 mg/kg, followed by maintenance doses of 90 mg every 8–12 weeks.
The Mayo Score for UC patients, the Crohn’s disease activity index for patients with Crohn’s disease, the levels of CRP, hemoglobin (Hb), and fecal calprotectin, as well as endoscopic and radiographic indicators of disease activity, were used to evaluate the medical activity of disease at the start of therapy.
During follow-up
The response to therapy was evaluated using the Mayo score [20], Crohn’s Disease Activity Index (CDAI) [21], and markers of inflammation, such as CBC and CRP concentrations, fecal calprotectin, and endoscopic and radiologic signs of disease activity.
Clinical response, remission, and laboratory investigations were evaluated 3, 6, and 12 months after starting therapy. At 6 months after the initiation of treatment, patients with UC were evaluated for endoscopic mucosal healing and histological remission, and patients with CD were evaluated using bowel ultrasound or magnetic resonance enterography.
Because of its affordable price, lack of side effects, and current accessibility, intestinal ultrasound has been a promising bedside technique for tracing clinical activity in IBD patients [22]. It is becoming more widely accepted as a valuable diagnostic tool for IBD, as well as for determining the severity of the inflammation, and early detection of complications, related either to the disease or to the therapeutic protocol, with results equivalent to other validated techniques. All examinations were performed by a single experienced (trained for several bowel ultrasound examinations under supervision of an ultrasound gastroenterologist specialist at Sacco Hospital, Italy) operator blinded to the clinical, laboratory, and endoscopic data of the patients. Patients were examined via ultrasound after a 6-hour fasting period to minimize intestinal air contents. The examination was performed using an ultrasound machine (Toshiba Xario, Japan) with a low frequency curved-array transducer (2.5–4.5 MHz) to determine any pathological bowel motility or distension and any para-intestinal structures, such as abscesses, in all abdominal quadrants. Examination with a high-frequency linear-array transducer (6.0–8.4 MHz) was used for bowel wall examination starting with examination of the proximal colon followed by the distal one and then the small bowel [23].
Statistical analysis
All analyzed data were revised, coded, tabulated, and introduced using SPSS Statistics, version 25.
Continuous variables were expressed as mean or median, while binary variables were expressed as numbers and percentages (%). The paired samples were compared by either the paired t-test or Wilcoxon signed-rank test. Fisher’s exact test was used to examine the relationship between two qualitative variables. The McNemar test was used to assess the statistical significance of the difference between qualitative variables measured twice for the same study group. A p-value less than 0.05 is considered significant.
Results
Descriptive data of UC patients
Among 16 patients, there were 6 males and 10 females; their ages ranged from 29 to 50 years, with a mean of 35 years. All were non-smokers with an average disease duration of 5 years. Before inclusion in our study, 87.5% received mesalazine orally, and 12.5% received topical mesalazine with or without oral form. All of them received oral steroids during their disease and 87.5% received intravenous (IV) steroids. In addition, all of them received azathioprine, 12.5% methotrexate, and 12.5% calcineurin inhibitors (CNIs) as rescue therapy.
Subsequently, all of them were steroid dependent and were shifted to biologic therapy (anti-TNF as the first line), 37.5% received infliximab, 25% received adalimumab, and 37.5% received both of them (one after failure of the other).
They were maintained on anti-TNF for an average of 1 year, when 62.5% had not responded (primary non-response), 25% responded initially then failed (secondary failure), and 12.5% partially improved.
They were subsequently scheduled to receive ustekinumab (75% of them), and the rest received vedolizumab.
Before starting treatment, all of them complained of bloody diarrhea, 75% had diarrhea, 87.5% had abdominal pain, 37.5% had bleeding per rectum, 62.5% had extraintestinal manifestations, and 50% had weight loss. Their Mayo score ranged from 5 to 9, with a median of 7.
Follow-up of UC patients after starting the second line
At first follow-up after 3 months
37.5% of them still complained of abdominal pain and diarrhea; bloody diarrhea partially improved, with a median of 4 motions per day, comparing laboratory data to that before starting the second line; the mean white blood cell (WBC) count dropped from 8.5 to 6.5/µl, Hb level rose from a mean of 9.9 to 10.6 g/dl, CRP dropped from a median of 8.8 to 4.35 mg/dl, red blood cells (RBCs) in stool analysis were still present in 12.5% of patients in comparison to 50% at the start, and presence of pus cells in stool dropped from 37.5% to 25% of patients, as shown in Table I.
Table I
Laboratory changes before and during treatment in UC receiving ustekinumab and vedolizumab
At the second follow-up at 6 months
37.5% of patients still complained of abdominal pain. The median (IQR) number of motions dropped from 4 (2–9) to 2.5 (2–7) motions/day. 14.3% still complained of extraintestinal manifestations in comparison to 62.5% before treatment. Hb level continued to rise (mean ± SD: 11.20 ±2.22 g/dl), CRP was median (IQR) 5.45 (4.00–13.10) mg/dl, and the Mayo score was median (IQR) 3.5 (2.5–7.5), as shown in Tables I and II.
Table II
Colonoscopy macroscopic changes before starting treatment in comparison to after 6 months of treatment in UC patients
Descriptive data of CD patients
Among CD patients, there were 4 males and 6 females, with a mean age of 34.00 ±10.97 years, 40% were smokers, with mean disease duration of 7.40 ±3.81 years. Before inclusion in our study, their main complaints were abdominal pain (100%), diarrhea (80%), fistulous opening (60%), and weight loss (60%). Previously, all of them received oral steroids, 80% received IV steroids, 80% received azathioprine, and 20% received methotrexate. Subsequently, 60% were steroid dependent while 40% were steroid-resistant; they were then shifted to anti-TNF biological treatment, with 60% receiving infliximab, 20% receiving adalimumab, and 20% receiving both (one after failure of the other) for a mean duration of 1.10 ±0.52 years.
While on anti-TNF, 60% had a primary non-response and 40% showed partial improvement. They were subsequently shifted to second-line treatment; 80% received vedolizumab and 20% received Ustekinumab.
Follow-up of CD patients after starting the second line
At the first follow-up after 3 months
All of them still complained of abdominal pain, WBC rose from 7.2 to 10.1/µl, CRP from 5 to 12 mg/dl, Hb slightly dropped from 11.38 to 10.6 g/dl. Also, stool analysis showed that 80% of them had RBCs and pus cells which were previously absent, as shown in Table III.
Table III
Laboratory changes before and during treatment in CD receiving ustekinumab and vedolizumab
At the second follow-up at 6 months
50% of those maintaining treatment still had abdominal pain; the average number of motions dropped from median (IQR) 10 (2–12) motions per day at 3 months of treatment to median (IQR) 4 (1.5–7) motions per day. 50% of those having fistulous openings did not improve, and the others partially improved. Thereafter, 80% of the patients continued treatment, whereas 20% stopped IT.
A colonoscopy was done at 6 months and showed no significant changes in comparison to the index colonoscopy except for some changes in the site of disease involvement, as shown in Table IV.
At the third follow-up after 1 year of treatment
After 1 year of treatment, 50% continued to complain of abdominal pain; the number of motions per day dropped to median (IQR) 2.5 (1–6.5) from the median (IQR) 4 (1.5–7) motions per day at the sixth month, and 75% showed no improvement in the fistulous opening, while 25% showed partial improvement. CDAI continued to drop to a median (IQR) of 124.5 (114–230). All data are shown in Table V.
Table V
Bowel ultrasound and CDAI changes before starting 2nd line and after 6 months of treatment in CD patients
Discussion
Only a few Egyptian studies have compared second-line drugs in IBD treatment. Choosing the best option for patients with moderate to severe UC or CD who have not responded to first-line anti-TNF therapy is crucial, based on the effectiveness and safety of the second-line agents that are currently on the market. This study focused on comparing the effectiveness of vedolizumab and ustekinumab in patients with UC and CD who received second-line drugs due to anti-TNF therapy refractoriness or intolerance.
From 50 IBD patients in our unit receiving biological therapy, 26 patients with severe UC and CD receiving vedolizumab and ustekinumab for 1 year were chosen for this study. These findings align with an Egyptian study which stated that IBD was a rare disease, and it additionally warned that the East Mediterranean countries should be aware of its presence [24].
Our study excluded patients exposed to more than one anti-TNF agent. However, other studies did not exclude these patients to study predictors of response to biological agents. For instance, an English non-randomized study showed that among patients who had received more than one anti-TNF medication, it had a negative impact on outcomes [25].
In the present study, 60% of CD patients had a primary non-response to anti-TNF and 40% showed partial improvement. However, in a French study, 21% of CD patients experienced a primary non-response to the first anti-TNF medication, which led to its discontinuation [26].
80% of CD patients in this study received vedolizumab, and 20% received ustekinumab. This was in contrast to a French multicenter study in which 44% of CD patients received ustekinumab, 21% received vedolizumab, and 35% received a second anti-TNF agent, while 75% of UC patients received ustekinumab, and 25% received vedolizumab.
UC patients received CNI as rescue therapy before starting second-line drugs in this study. Ollech et al. describe a single-center experience with CNIs to induce a therapeutic response followed by vedolizumab maintenance therapy in patients with steroid-refractory UC. This retrospective study adds significantly to the value of this treatment approach [27].
During follow-up, none of the patients died or were hospitalized as a result of the study drugs’ side effects. There were 3 patients with severe headaches in the ustekinumab group. This is consistent with Biemans et al., who found that there were no adverse events, infection rates, or hospitalizations in the vedolizumab and ustekinumab treatment groups [28]. It contrasts with an English retrospective study that reported the death of 2 patients during the follow-up period, but neither was directly related to the treatment [25].
In this study, the majority of CD patients receiving vedolizumab experienced longer disease duration than UC patients, and were also older than in the UC group. This contrasted with a Korean study in which the UC patients receiving vedolizumab were older, with longer disease duration [29].
This reflects the well-known safety of vedolizumab in elderly patients as well as the fact that older patients frequently have longer disease durations [29]. Longer disease duration was previously shown to be associated with negative outcomes with vedolizumab [12].
The majority of CD patients (80%) received vedolizumab, while the majority of UC patients (75%) received ustekinumab. UC activity as indicated by the Mayo Scoring Index significantly decreased during follow-up. This supports the finding that second-line drugs with mechanisms other than those of anti-TNF agents can be used to achieve a more effective response.
However, CD activity as indicated by CDAI did not significantly decrease together with the rise of inflammatory marker levels. This result confirms ustekinumab’s superior effectiveness for treating anti-TNF refractory patients over vedolizumab and suggests a lower response rate after vedolizumab administration.
This is consistent with a retrospective French study that reported superior efficacy for ustekinumab at week 48 in achieving clinical remission [30]. In keeping with this, another Dutch study also confirmed the superior effectiveness of ustekinumab compared to vedolizumab for treating anti-TNF refractory patients [26]. To the contrary, another Dutch study showed comparable effectiveness between vedolizumab and ustekinumab in treating anti-TNF refractory CD [31].
In patients with CD, CRP was elevated during follow-up. However, several studies were unable to confirm associations with clinically active disease or an elevated CRP and response to ustekinumab or vedolizumab [25].
One of our study limitations is that the number of enrolled patients is small, and variations in study design could result in inconsistent results. Head-to-head studies with larger numbers, especially in CD patients with prior anti-TNF treatment failure, are needed. Future research will require extensive prospective studies, particularly in Egypt.
Conclusions
This study demonstrated that, in bioexperienced IBD patients – especially those with UC – the outcome was better with ustekinumab than vedolizumab use as a second line of treatment among UC patients in whom previous anti-TNF treatment had failed. In the meantime, our findings may help guide clinical decision-making in IBD patients following anti-TNF failure.
