Efficacy, Effectiveness, and Safety of Ustekinumab for the Treatment of Ulcerative Colitis: A Systematic Review

Abstract

Objective

This study aimed to evaluate ustekinumab efficacy, effectiveness, and safety as a treatment for ulcerative colitis in adult patients.

Methods

A systematic review of the efficacy, effectiveness, and safety of ustekinumab in ulcerative colitis was carried out. The search was conducted via PubMed, Embase, and the Cochrane library. Two reviewers independently assessed the quality of studies and extracted study data.

Results

Of the 892 studies identified, 17 were included: 1 randomized controlled trial (RCT), 3 long-term extensions, and 13 observational studies. In the randomized clinical trial evaluating efficacy at week 8, clinical remission was achieved in 16% of patients, whereas clinical response was achieved in 51% and 62% of patients who received intravenous ustekinumab at a dose of 130 mg and 6 mg/kg, respectively. At 3 years’ follow-up, symptomatic remission was achieved in 68% of patients. On the other hand, the effectiveness of ustekinumab was evaluated in 13 observational studies. In these studies, clinical remission at induction was achieved in 24% to 61% of cases, whereas clinical response at induction was achieved in 47% to 77% of cases. Moreover, clinical remission was achieved in 33% to 79% of cases at 52 weeks of follow-up. The adverse events ranged from 2.6% to 77% of all the studies that reported safety data. Adverse events leading to discontinuation ranged from 2.6% to 8.1%, and serious adverse events were uncommon and ranged from 3.7% to 6.0%.

Conclusions

Ustekinumab has demonstrated efficacy (in RCTs), effectiveness (in real clinical practice), and safety for the treatment of ulcerative colitis.

Key messages.

• What is already known?

Ustekinumab has demonstrated to be effective and safe for the treatment of ulcerative colitis in 1 randomized control trial.

• What is new here?

The main strength is that this review included both randomized clinical trials (including long-term extension) and observational studies assessing ustekinumab for the treatment of ulcerative colitis.

• How can this study help patient care?

Due to a large number of clinical trials and observational studies published, clinicians may rely on available systematic reviews to obtain information on the evidence regarding a specific drug. Therefore, the present systematic review on ustekinumab may be used to guide decision-making.

Introduction

Ulcerative colitis (UC) is a chronic disease characterized by inflammation of the colon with a relapsing/remitting course. Clinical manifestations vary according to the extent of the disease and symptoms.¹ The worldwide incidence and prevalence of UC have increased over time. In Europe, the incidence has been estimated at 24 per 100 000 person years.¹ The prevalence rate has been estimated at approximately 80 cases per 100 000 inhabitants,² and mortality has been estimated at 0.50 per 100 000.²

Ustekinumab is a fully human IgG1 monoclonal antibody to interleukin 12/23 currently approved for the treatment of adult patients with moderately to severely active UC who have had an inadequate response with, lost response to, or were intolerant to either conventional therapy or a biologic, or have medical contraindications to such therapies.³

Ustekinumab has been proven effective and safe for the induction and maintenance of UC remission, according to the UNIFI clinical trial program⁴ and the long-term Extension study.⁵ Nevertheless, due to the strict inclusion criteria used in randomized controlled trials (RCTs), their results may not be representative of standard clinical practice in UC, as previous studies have shown that only 26% of patients with UC would have been eligible to participate in selected RCTs. Age, disease severity, and comorbidities are among the exclusion criteria most commonly used in RCTs regarding inflammatory bowel disease.^6,7 Accordingly, studies evaluating the effectiveness and safety of ustekinumab for UC in routine clinical practice may provide additional information to that reported in RCTs.^8–10 Therefore, the purpose of this systematic review was to evaluate the efficacy (from RCTs), the effectiveness (from observational studies), and the safety of ustekinumab for the treatment of adult patients with UC.

Methods

This study was performed following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.¹¹ The protocol of the systematic review was registered in the international prospective register of systematic reviews (PROSPERO) on November 1, 2021 (ID number CRD42021282254).

Eligibility Criteria

This systematic review included studies published in English that evaluated the efficacy, effectiveness, and/or safety of ustekinumab for the treatment of adult patients with UC. No restrictions were applied regarding the date of publication. Studies whose purpose was outside the scope of this review, studies performed on pediatric populations, editorials, commentaries, letters, personal narratives, clinical practice guidelines, and technical reports were excluded.

Sources of Information

The search strategy was implemented in the following electronic databases: PubMed, Embase, and the Cochrane Library. In addition, conference proceedings from the American Gastroenterological Association (AGA), United European Gastroenterology (UEG), and the European Crohn’s and Colitis Organisation (ECCO) were searched.

Search Strategy

The search strategy was adapted to each database and performed on March 16, 2022 (Supplementary Table S1).

Selection Process

Initially, duplicate studies were identified and excluded. Thereafter, 2 reviewers independently screened the studies according to the established inclusion and exclusion criteria. In case of disagreement, a third reviewer assessed the eligibility of the article. Studies were peer-reviewed by title and abstract, followed by the full-text of those considered potentially eligible, indicating the reason for exclusion.

Data Collection Process

Data extraction from the selected studies was performed by 2 reviewers. The following information was recorded: general characteristics of the publication (eg, identification number, year of publication, author, and whether it was a full article or an abstract from a conference proceeding); study design; study population (eg, sample size, age, sex, duration of disease, and UC classification/extension); treatment (eg, previous biologic treatment, induction or maintenance, and dose); efficacy (eg, clinical remission, clinical response, endoscopic remission, and endoscopic response); safety (eg, adverse events [AE], serious AE [SAE], infections, and malignancies), colectomy; and persistence of the drug (discontinuation and reasons for discontinuation).

Risk of Bias of Included Studies

The risk of bias in the selected studies was independently assessed by 2 reviewers. In case of disagreement, a third reviewer evaluated the study for its risk of bias. Randomized control trials were assessed using the Cochrane risk of bias tool (RoB-2).¹² This tool assessed bias in 5 distinct domains to classify studies as having “low risk of bias,” “some concerns,” or “high risk of bias.” Moreover, observational studies were assessed using the Newcastle-Ottawa quality assessment scale, with 9 being the highest obtainable score¹³ and 7 or more being a high-quality score.

Data Synthesis and Presentation

A descriptive analysis of the most relevant variables was performed to summarize the results of this systematic review. The mean (standard deviation) or median (interquartile range) were extracted from studies to describe patient age and disease duration. The percentage of patients achieving response or remission in the different treatment phases and the percentage of AEs were further extracted from each study. Dispersion plots were generated to provide a visual summary of efficacy over time.

Results

Selection of Studies

The initial search yielded 892 records (232 in PubMed, 542 in Embase, 118 in the Cochrane Library, and 38 relevant abstracts from conference proceedings), which were reduced to 523 articles after removing duplicates. Once reviewed by title and abstract, 60 articles or abstracts from conference proceedings met the inclusion criteria. Finally, after a full-text review, 17 studies were selected to be included in this systematic review.^{4,5,8–10,14–25}Supplementary Figure S1 shows the flow chart of this systematic review.

General Characteristics of Studies and Population

Most of the included studies were full-text articles (59%) and were published in 2021 (59%). Eight studies were part of a clinical program or registry, the most frequent being the UNIFI program. Moreover, 76% were observational studies. Of the studies that evaluated efficacy or effectiveness, only 1 retrospective observational study used an active comparator (tofacitinib),²² whereas the remaining studies used a placebo or did not indicate the study comparator. Table 1 describes the characteristics of the included studies.

Table 1.

General characteristics of the studies included.

	Number of studies	%
Publication
Full-text articles	10	59
Congress abstracts	7	41
Publication year
2019	1	5.9
2020	2	12
2021	10	59
2022	4	24
Clinical study/ register
UNIFI	4	24
GETAID	2	12
ENEIDA	1	5.9
SWIBREG	1	5.9
Not specified	9	53
Study design
Randomized controlled trial (RCT)	1	5.9
Extension study-RCT	3	18
Observational study	13	76
Clinical outcomes reported
Efficacy (RCT)	4	24
Effectiveness (observational study)	13	76
Comparator
Placebo	2	12
Tofacitinib	1	5.9
None	14	82

Abbreviations: RCT, Randomized controlled trial.

Regarding patient characteristics of the included studies, the percentage of men ranged from 35% to 69%, and the median age ranged from 27 to 46 years. The median disease duration was between 5 and 9 years. Of the studies that reported data on UC extension, approximately half of the patients had extensive colitis, except for 1 study²³ that reported only 26%. The Supplementary Material shows the characteristics of included studies in detail (Table S2).

Risk of Bias of Included Studies

Of the 4 RCTs evaluated, 1 had a low risk of bias,⁴ 2 were rated as having some concerns,^5,16 and 1 had a high risk of bias.²⁵ Regarding observational studies, while none of the included studies had high quality (a score of 7 or more), those of Chaparro et al. (2021)⁹ and Dalal et al. (2021)²² had the highest quality (a score of 6 out of 9), indicating a lower risk of bias than the remaining observational studies. On the contrary, there were 2 studies with a significantly low-quality rating (a score of 2 out of 9), indicating a large risk of bias.^20,23 The results of the quality assessments are presented in Tables 2 and 3 for RCTs and observational studies, respectively. Furthermore, the Supplementary Material shows the results of the risk of bias assessment in detail (Tables S3 and S4).

Table 2.

Risk of bias in randomized clinical trials.

Reference (Author, Year)	Risk of Bias Domains (D)
	D1	D2	D3	D4	D5	Overall
Abreu, 20225
Leong, 202225
Sands, 20194
Scherl, 202116

Low risk.

Some concerns.

High risk.

Domains:

D1 Bias arising from the randomisation process.

D2 Bias due to deviations from intended interventions.

D3 Bias due to missing outcome data.

D4 Bias in measurement of the outcome.

D5 Bias in selection of the reported result.

Table 3.

Risk of bias in observational studies using the Newcastle-Ottawa score.^a

Reference	Selection	Comparability	Outcome	Overall
Alsoud, 202223	*		*	**
Amiot, 20208	**		**	****
Chaparro, 20219	***		***	******
Chiappetta, 202118	**		***	*****
Dalal, 202110	**		**	****
Dalal, 202122	***	*	**	******
Ecker, 202121
Fumery, 202119	**		**	****
Hong, 202114	**		**	****
Iborra, 202224			**	**
Ochsenkühn, 202115	**		**	****
Deepak, 202120	*		*	**
Thunberg, 202117	**		**	****

^aMaximum: 9 points. A score of ≥7 was suggestive of a high-quality study.

Efficacy of Ustekinumab (RCTs: UNIFI)

After induction (at week 8), both ustekinumab treatment arms (130 mg and 6 mg/kg) achieved statistical significance vs placebo for clinical remission, clinical response, endoscopic improvement, or histo-endoscopic mucosal healing.⁴ At week 8, the percentage of patients that met primary and secondary end points were significantly higher in both ustekinumab treatment arms compared with the placebo group, 16% of patients achieved clinical remission at week 8 in both ustekinumab treatment arms,⁴ whereas clinical response was achieved by 51% and 62% of patients who received 130 mg and 6 mg/kg, respectively. At week 8, endoscopic improvement was achieved by 26% and 27% of patients when treated with 130 mg and 6 mg/kg, respectively, and histo-endoscopic mucosal healing was achieved in 18% (ustekinumab 130 mg) and 20% (ustekinumab 6mg/kg) of patients at week 8, with a significant statistical difference compared with placebo (P < .001).⁴ Efficacy was maintained over time. At week 44, the percentage of patients who achieved clinical remission was significantly higher among those assigned to 90 mg of subcutaneous ustekinumab every 12 weeks (38%) and every 8 weeks (44%) compared with placebo (24%).⁴ Similarly, endoscopic improvement was significantly higher in patients receiving ustekinumab (44% of patients treated every 12 weeks and 51% of patients every 8 weeks) compared with placebo (29%).⁴ Abreu et al⁵ recently published results from the long-term extension of UNIFI; at years of follow-up, 66% and 69% of patients receiving ustekinumab every 12 and 8 weeks, respectively, were in symptomatic remission.⁵ Through week 152, biologic-naïve patients were those with the highest proportion of patients in symptomatic remission.⁵ Discontinuation of ustekinumab due to a lack of efficacy at 3 years in patients entering the long-term extension was reported in 4.2% of cases.⁵ In a subanalysis of the UNIFI study, Leong et al²⁵ showed that there was a higher proportion of patients in symptomatic remission in those who achieved histo-endoscopic mucosal healing after induction compared with those with only histological or only endoscopic improvement. Table 4 shows the detailed efficacy data of included studies.

Table 4.

Efficacy of ustekinumab in ulcerative colitis (RCTs).

Author, Year	N Patients	Design	Biologic Experience (%)	Induction				Maintenance		Follow-up (months)	Discontinua-tion n (%)	Reasons for Discontinuation
				Clinical Remission (%)	Clinical Response (%)	Endoscopic Improvement (%)	Predictors of Short-Term Response	Clinical Remission (%)	Predictors of Long-term Response
Sands et al, 20194	961: 642 UST 319 placebo	RCT	Treatment failure with biologics: 51%a and 52 %b	16% (50/320)a 16% (50/322)b	51% (164/320) a 62% (199/322) b	26% (84/320)a 27% (87/322)b	NA	38% (66/172)c 44% (77/176)d	NA	9.8	NA	NA
Abreu et al, 20225*	284 UST 115 placebo	LTE	Overall, 44% of patients had a history of biologic failure	NA	NA	NA	NA	66% (93/141)e 69% (99/143)f 74% (61/82)g 81% (54/67)h 55% (29/53)i 58% (41/71)j	NA	36	20% (57/284) 4.2% (12/284) 1.1% (3/284) 6.7% (19/284)	All causes Lack of efficacy Did not show improvement in UC activity 16 weeks after dose adjustment Other
Leong et al, 202225*	438	LTE	NA	NA	NA	NA	NA	70% (81/116)k 47% (14/30)l 55% (58/106)m	NA	36	NA	NA

Abbreviations: LTE, long-term extension; NA, nonavailable; RCT, randomized controlled trial; UST, ustekinumab

*Symptomatic remission.

^aUST 130 mg

^bUST 6 mg/kg

^cUST 90 mg q12w

^dUST 90 mg q8w

^eUST 90mg q12w or q8w after dose adjustment overall population.\

^fUST 90mg q8w overall population

^gUST 90mg q12w or q8w after dose adjustment Biologic naïve patients

^hUST 90mg q8w Biologic naïve patients

ⁱUST 90mg q12w or q8w after dose adjustment Biologic therapy failure

^jUST 90mg q8w Biologic therapy failure

^kPatients with histo-endoscopic mucosal healing at week 8

^lPatients with endoscopic improvement but no histologic improvement at week 8

^mPatients with histologic improvement but no endoscopic improvement at week 8

Effectiveness of Ustekinumab:Observational Studie

Thirteen studies reported effectiveness results and are summarized in Table 5. Clinical remission after 4 to 16 weeks of induction was reported in 7 studies and ranged from 24% to 61%. Steroid -free remission was reported in 6 studies and ranged from 14% to 67%. In most studies that reported clinical remission in induction and maintenance, the percentage of patients achieving remission increased at 24 or 52 weeks. At 6 months of ustekinumab treatment, clinical remission was observed in 40% to 58% (n = 8 studies) of patients and in 33% to 79% at 12 months. Moreover, clinical response was achieved in 47% to 77% of patients after induction, as reported in 8 different studies. Endoscopic remission was achieved in 21% to 83% of patients (n = 3 studies), and endoscopic response was achieved in 52% to 56% of patients (n = 2 studies). Lack of effectiveness, refractory disease, and loss of response were the main reasons for ustekinumab discontinuation (Table 5). Finally, histo-endoscopic mucosal healing was reported in 1 study by Hong et al,¹⁴ with 33% of patients (4 of 12) achieving this end point at 12 months. There was some heterogeneity in outcome definition that may explain the variability of results; a detailed definition of outcomes for each study is provided in the Supplementary Material (Table S5).

Table 5.

Effectiveness of ustekinumab in ulcerative colitis (observational studies).

Author, Year	N Patients	Design	Biologic Experience (%)	Induction					Maintenance		Follow-up (months)	Discontinuation (%)	Reasons for Discontinuation
				Clinical Remission (%)	Clinical Response (%)	Endoscopic Remission (%)	Endoscopic Response (%)	Predictors of Remission/Response	Patients in Clinical Remission (%)	Predictors of Long-term clinical Remission
Hong et al, 202114	66	Retrospective study	Prior use of any biologic or novel small molecule: 92%	43% (20/47)	49% (23/47)	NA		History of TNF-antagonist primary nonresponse and baseline Mayo endoscopic score of 3 were significant negative predictors of clinical remission at 3 months	45% (9/20)		5.9	30% (8/27) after dose escalation	NA
Ochsenkuhn et al, 202015	19	Retrospective study	All patients had been treated with anti-TNF-antibodies and anti-integrin-antibodies	37% (7/19) month 1, 58% (11/19) month 3	NA			NA	58% (11/19) month 6, 53% (10/19) month 9, 53% (10/19) month 12	NA	12	21% (4/19)	Refractory disease
Thunberg et al, 202117	145	Multicenter study	Previous anti-TNF-antibodies ≥ 1: 94%	NA	NA	NA		NA	NA	NA	7.4	NA	NA
Amiot et al, 20208	103	Multicenter retrospective study	Anti-TNF therapy: 99%	40% (41/103)	53 (55/103)	NA		NA	NA	NA	2.8 – 3.7	1.9% (2/103) before the week 12-16 9.7% (10/103) at the week 12-16 visit,	Lack of efficacy Lack of efficacy
Chaparro et al, 20219	95	Multicenter prospective study	Prior biological treatment or tofacitinib: 100%	35% (33/95)	53 (50/95)	NA		CRP serum concentration over the normal range upper limit was the only factor significantly associated with lower probability of achieving remission	33% (18/54)	NA	19	22% (21/95) 13% (12/95)	Primary nonresponse Loss of response Neither
Chiappetta et al, 202118	68	Multicenter retrospective study	Previous biologics: 97%	NR	62% (42/68)	NA		NR	NR	NA	12	1.5% (1/68) week 8 1.6% (1/61) week 24 16% (6/38) week 52	Primary failure Secondary failure Secondary failure
Fumery et al, 202119	103	Multi-center retrospective study	Anti-TNF therapy: 99%	NA	NA	NA		NA	34% (35/103)	NA	12	44% (41/103) 1% (1/103) 1% (1/103)	Lack of effectiveness Pregnancy Personal decision
Deepakl et al, 202120	95	Multicenter study	Only 4.3% were naïve to biologics or tofacitinib and 62 (66%) had previous exposure to at least 2 other biologics	24%	47%	21% (7/33)		No variables were found to be associated with response at week 8	40%	Obesity (BMI ≥ 30)	5	NA	NA
Ecker et al, 202121	26	Retrospective study	Only 4% were anti-TNF naïve	NA	77%	NA		NA	NA	NA	26	7.7% (2/26) 3.8% (1/26)	Lack of improvement of symptoms Colectomy
Dalal et al, 202110	108	Retrospective study	91.7% had prior anti–TNF exposure	NA	68% (27/40)		56% (23/40)	NA	NA	NA	7.7	30% (12/40): 10/12 2/12	Had no/minimal response to induction Had loss of response.
Alsoud et al, 202223	42	Prospective study	Previous biological/ small molecules: 90%	57% (24/42)	74% (31/42)	26% (11/42)	Improvement: 52% (22/42)	NA	NA	NA	6	NA	NA
Iborra et al, 202224	23		All patients had received previous biological treatment	61% (14/23) at 8w 58% (13/23) at 16w 56% (13/23) at 24w		83% (5/6)		NA	79% (18/23) 52w	NA	12	13% (3/23)	NA

BMI: Body mass index; NA: nonavailable; TNF: tumour necrosis factor.

Five observational studies^{9,14,15,20,24} assessed clinical remission at 2 or more time periods (Supplementary Figure S2). At week 24, a rise in clinical remission was observed in 2 studies (69% and 11% from baseline at week 8 and 16, respectively). Two other observational studies also showed an increase in clinical remission at week 52 (from baseline). The proportion of patients achieving clinical response raised at weeks 24 and 52 from baseline (an increase from 11% to 36%; Supplementary Figure S3). The proportion of patients achieving corticosteroid-free clinical remission increased at weeks 24 and 52 compared with baseline (in 4 out of 5 observational studies reporting this variable at 2 or more time periods; Supplementary Figure S4).

Predictive Factors of Clinical Remission/Response

The included RCTs reported no predictors of remission or response to ustekinumab treatment. Negative predictors of clinical remission at 3 months in observational studies were a history of tumor necrosis factor (TNF)-antagonist primary nonresponse and a baseline Mayo endoscopic score of 3.¹⁴ Chaparro et al⁹ showed that C-reactive protein (CRP) serum concentration over the upper limit of the normal range was the only factor significantly associated with a lower probability of achieving remission. Deepak et al²⁰ identified that the rate of response was lower in obese patients (body mass index ≥30). Furthermore, the Supplementary material shows the results of the predictive factors detail (Tables S6).

Safety Outcomes

Table 6 shows the safety profile of ustekinumab. Two RCTs studies^4,5 and 11 observational studies^{8–10,14,15,18–22,24} reported safety results. The percentage of patients with AEs ranged from 41% to 77% in RCTs, whereas in observational studies patients with AEs ranged from 2.6% to 32%.¹⁵ Moreover, patients with SAEs were uncommon and ranged from 3.7% to 6%. Infections were reported in 1.9% to 49% of patients; however, serious infections ranged from 0.3% to 3.5%. Patients with AEs leading to discontinuation ranged from 2.6% to 8.1%. Few studies reported malignancies, and these ranged from 0.6% to 5.3%. Colectomy was reported in 3.8% to 11% of patients. Only 1 study⁵ reported patient-years (PYs) of follow-up. According to this study,⁵ from maintenance week 0 to 156, patients who received ustekinumab (every 12 or 8 weeks combined) had 1236 PYs of follow-up.

Table 6.

Safety of ustekinumab.

Author, year	N patients	Adverse events	Serious adverse events	Adverse events leading to discontinuation	Infections	Serious infections	Malignancies	Colectomy
Sands et al, 20194	961	41% (133/321)a 51% (162/320)b 69% (119/172)c 77% (136/176)d	3.7% (12/321)a 3.4% (11/320) b 7.6% (13/172)c 8.5% (15/176)d	5.2% (9/172)c 2.8% (5/176)d	16% (51/321)a 16% (51/320)b 34% (58/172)c 49% (86/176)d	0.6% (2/321)a 0.3% (1/320) b 3.5% (6/172)c 1.7% (3/176)d	0.6% (1/172)c 0.6% (1/176)d	NA
Abreu et al, 20225	457	251.90 (241.64, 262.49 e	8.0 (6.51, 9.75)e	8.1 (23/284)	76.11 (71.32, 81.13) e	2.43 (1.64, 3.46)e	0.73 (0.33, 1.38) e	NA
Hong et al 202114	66	12% (8/66)	6.0% (4/66)	NA	6.1% (4/66)	NA	0	4.5% (3/66)
Ochsenkuhn et al 202015	19	32% (6/19)	0	5.3% (1/19)	11% (2/19)	0	5.3% (1/19)	11% (2/19)
Amiot et al 20208	103	7.8% (8/103)	3.9% (4/103)	NA	1.9% (2/103)	NA	0	NA
Chaparro et al 20219	95	3.2% (3/95)	NA	1.1% (1/95)	2.1% (2/95)	1.1% (1/95)f	NA	9.5% (9/95)
Chiappetta et al 202118	68	2.6% (1/38)	0	2.6% (1/38)	NA	NA	NA	NA
Fumery et al 202119	103	15 % (15/103)	3.9% (4/103)	3.9% (4/103)	6.8% (7/103)	NA	0	9.7% (10/103)
Deepak et al 202120	95	6.6% (6/95)	0	0	NA	NA	0	5.3% (5/95)
Ecker et al, 202121	26	NA	NA	0	3.8% (1/26)	NA	3.8% (1/26)	3.8% (1/26)
Dalal et al, 202110	108	5% (2/40)	NA	NA	5% (2/40)	NA	NA	NA
Dalal et al,202122	36	5.6% (2/36)	NA	NA	NA	NA	NA	NA
Iborra et al,202224	23	29% (2/23)	NA	NA	NA	NA	NA	4.3% (1/23)

NA: nonavailable.

^austekinumab 130 mg.

^bustekinumab 6 mg/kg.

^custekinumab 90 g q12w.

^dustekinumab 90 g q8w.

^eNumber of events per 100 patient-years of follow-up (95% CI).

^fsevere SARS-CoV-2 infection.

Discussion

This systematic review estimates the short-term and mid-term efficacy of ustekinumab to induce and maintain remission in patients with UC, which were 16% and 41%, respectively. The long-term efficacy was further demonstrated, as symptomatic remission was achieved in 68% of patients at year 3 of follow-up. These results have been confirmed by observational studies evaluating ustekinumab in routine clinical practice, which showed that approximately 24% to 61% of patients achieved clinical remission after induction; this ranged between 33% and 79% at 52 weeks.

The percentage of patients with endoscopic remission at induction was similar in the RCTs and in 2 of the observational studies that reported it.^20,23 In another observational study,²⁴ this percentage was much higher, possibly due to its smaller sample size (23 patients). Another similarity between efficacy results (from RCTs) and effectiveness results (from observational studies) was clinical remission in maintenance, with the percentage of patients achieving remission ranging from 38% to 68% in RCTs and 33% to 79% in observational studies.

On the other hand, differences were observed in the percentage of patients achieving clinical remission, clinical response, and endoscopic remission in short-term RCTs and observational studies, which were higher in observational studies. These differences are probably due, among other reasons, to differences in inclusion criteria (more restrictive in RCTs), sample size (larger in RCTs), measurement time (although short-term outcomes were measured at 4, 8, 12, or 16 weeks), and study design (prospective in RCTs and retrospective in most observational studies).

In the observational studies included, the results of the outcome variables were not homogeneous. This may be due to differences in sample size (ranging from 19 to 145 patients), study design (retrospective and prospective), and study setting (multicenter or single center).

Notably, although no RCT reported predictors of clinical remission in this systematic review, 3 observational studies found significant predictors. Specifically, CRP, history of TNF-antagonist primary nonresponse, and higher baseline Mayo score were significantly associated with a lower probability of achieving remission. Additionally, obesity was associated with a lower response. Other possible predictor factors such as disease duration (≥5 years), pancolitis, combination therapy with immunomodulators, aminosalicylate, or previous exposure to other drugs (≥2 biologics, ≥3 biologics, vedolizumab, tofacitinib) have been studied, but no significant associations were reported.

In addition, a retrospective study²⁶ identified corticosteroid withdrawal in the first 3 months and clinical response at 6 months after treatment initiation as significant predictors of clinical remission in patients with active UC treated with anti-TNF. However, a systematic review²⁷ of potential predictors of a short-term primary response to biologic treatment concluded that, in the case of ustekinumab, no predictive factor has been yet reported to be helpful in routine clinical practice.

The safety profile of ustekinumab may be comparable to other treatments for UC; for example, SAEs were identified in 9.0% to 17% of patients treated with vedolizumab,²⁸ in 14% of those treated with adalimumab,²⁸ and in 4.0% of those treated with tofacitinib.²⁹ In our systematic review, SAEs were identified in 3.4% to 8.5% of patients treated with ustekinumab. The occurrence of serious infections was lower among patients treated with ustekinumab (3.5%) than among those treated with vedolizumab (19%).²⁸ In addition, like with vedolizumab,²⁸ most studies reporting on the advent of malignancy reported an incidence of less than 1% among patients treated with ustekinumab. Finally, treatment discontinuation due to AEs was identified in 1.1% to 8.1% of patients treated with ustekinumab, in 0% of those treated with vedolizumab,³⁰ and in 5% of those treated with adalimumab.³⁰ Notably, the study reporting 8.1% had a follow-up of 152 weeks.

In a recent network meta-analysis,³¹ no differences in AEs and SAEs were observed between active interventions. Nevertheless, ustekinumab ranked third (after vedolizumab and golimumab) in terms of safety when considering SAEs, and ozanimod ranked last (ie, greater occurrence of SAEs).³¹ The most common noninfectious AEs associated with ustekinumab, identified in a recent metanalysis,³² were arthralgia (57%), rash or pruritus (51%), and headache (46%). According to Honap et al,³² the infections associated with the use of ustekinumab were predominantly pulmonary (55%) or gastrointestinal (29%). Moreover, events of herpes zoster were reported in 13 of 17 studies associated with the use of tofacitinib.²⁹ In addition, sphingosine 1-phosphate modulators (such as ozanimod) were associated with an increased risk of herpes zoster infection and transient cardiovascular events.³³

There are several limitations in this study. First, the abstracts selected from specific conference proceedings reported insufficient information to assess their quality. Furthermore, only 1 RCT study had a low risk of bias, and no observational study had high quality. However, including the abstracts was considered appropriate given the scarce scientific evidence on the effectiveness of ustekinumab in UC compared with Crohn’s disease, and the relevance of the conferences considered. Second, most of the observational studies that evaluated effectiveness were retrospective, with inherent limitations such as selection bias and confounding. Only 1 of these studies used intervention and control groups, comparing the effectiveness of tofacitinib with ustekinumab; the other observational studies evaluated the variables of interest (effectiveness and/or safety) in a cohort of UC patients treated with ustekinumab, without a control group. Third, the small sample size of some of the included observational studies makes extrapolation of results difficult. In observational studies, the effectiveness depends on clinical subjective assessment, which leads to inter- and intraobserver heterogeneity and makes the comparison between studies difficult. Outcome definition and time point assessment may vary among different studies. Furthermore in observational studies, information on endoscopic healing is very limited and probably unreliable, because in clinical practice colonoscopy is performed precisely on those patients with more severe disease or who are less responsive to therapy. Although observational studies have their limitations, they are necessary to evaluate the results of an intervention under clinical practice conditions. However, they should be critically reviewed when interpreting their results. In this regard, this systematic review provides important information on the effectiveness and safety of ustekinumab in the treatment of UC in routine clinical practice.

On the other hand, our systematic review has several strengths. The main one is that it included both RCTs and observational studies assessing ustekinumab for the treatment of UC. Furthermore, previous systematic reviews have evaluated the effectiveness and safety of this drug; however, in contrast to the systematic review of effectiveness by Honap et al,³² which included 3 effectiveness studies in UC with 141 patients, the present systematic review included 13 studies with a total of 893 patients from observational studies. Moreover, another review focused only on the efficacy (ie, only RCTs) of ustekinumab for the treatment of UC, which may not be representative of routine clinical practice.³⁴ Another strength was that the present systematic review included both manuscripts in main search engines and articles presented at relevant conferences. We also assessed the methodological quality of the included studies. However, a detailed analysis of clinical variables was performed at various points in time throughout the follow-up of the included studies for both RCTs and cohort studies. Finally, we analyzed the tendency (increase or decrease) for each of the main clinical variables measured at different points in time and for each observational study.

In conclusion, ustekinumab has not only demonstrated efficacy in RCTs (and their extension studies) but also effectiveness and safety in observational studies from real clinical practice.

Funding

Project was funded by Janssen. However, the authors were independent in the design of the study, the analysis, and the discussion of the data.

Conflicts of Interest

J.G. has served as speaker, consultant, and advisory member for or has received research funding from MSD, Abbvie, Pfizer, Kern Pharma, Biogen, Mylan, Takeda, Janssen, Roche, Sandoz, Celgene, Gilead/Galapagos, Ferring, Faes Farma, Shire Pharmaceuticals, Dr. Falk Pharma, Tillotts Pharma, Chiesi, Casen Fleet, Gebro Pharma, Otsuka Pharmaceutical, and Vifor Pharma. M.C. has served as a speaker or has received research or education funding from MSD, Abbvie, Hospira, Pfizer, Takeda, Janssen, Ferring, Shire Pharmaceuticals, Dr. Falk Pharma, and Tillotts Pharma. E.P-R. is an employee of Weber, a consultancy firm that received fees from Janssen to develop this study.