Review article: the role of anti-TNF in the management of ulcerative colitis – past, present and future


  • This commissioned review article was subject to full peer-review.

Correspondence to:

Prof. S. Danese, IBD Center, IBD Center, Humanitas Clinical and Research Centre, 20089

Rozzano, Milan, Italy.




Until recently, the management of ulcerative colitis (UC) consisted of the stepwise use of mesalazine, corticosteroids and immunomodulators, or consideration of surgery. Anti-tumour necrosis factor (TNF) agents are recent additions to the UC-treatment algorithm.


To provide clinicians with a review of the role of anti-TNFs in UC, discussing how the drug(s) were used in the past, their current use and to determine their future role.


The scientific literature was reviewed to evaluate data on the use of anti-TNFs in UC.


In this review, we report how the management of UC has changed with the availability of anti-TNFs. The results from landmark anti-TNF trials have impacted clinical practice, leading to a readjustment of treatment goals. In addition, experience from clinical trials and local real-life cohorts have helped to clarify some misunderstandings in the management of UC. New anti-TNFs are on the horizon but questions still remain on the future role of anti-TNFs with regard to impact on disability, digestive damage and the possible development of risk matrices. Experiences from the use of anti-TNFs in Crohn's disease (for example, combination therapy and early treatment) now need to be addressed in UC.


The use of anti-TNFs in the management of UC has matured rapidly. Clinical experience has helped shape the current role of anti-TNFs, but more clinical research is needed to optimise their future role.


Over a decade ago, treatment goals in ulcerative colitis (UC) focused mainly on achieving symptomatic remission. Eliminating steroid use and aiming for endoscopic remission were not considered. The armamentarium for UC ranged from aminosalicylates for mild disease to IV steroids and calcineurin inhibitors for severe cases and treatment decisions were based mainly on disease severity.[1] Abundant evidence exists supporting the use of aminosalicylates [5-aminosalicylic acid (5-ASA)] in mild to moderate UC for inducing remission. Corticosteroids are used in patients not responding to 5-ASA or in patients with severe disease. Azathioprine (AZA) and mercaptopurine have been shown to be useful in steroid-refractory patients. Cyclosporin is mainly reserved for severe patients after IV corticosteroids.[1] A new class of drugs, anti-tumour necrosis factor (TNF) biologics, has emerged over a decade ago, and were originally evaluated in Crohn's disease (CD). More recently, the efficacy of these drugs was assessed in UC in controlled trials and local cohort experience demonstrating efficacy in moderate-to-severe UC.

This article will review how the introduction of anti-TNF agents has changed the management of UC over time, how these agents are currently used and what clinicians have to look forward to in the future with regard to managing UC with anti-TNFs.

Anti-TNF use in the past

In the early 2000s, anti-TNF agents were being evaluated for efficacy and safety in UC. Infliximab is a monoclonal antibody that binds TNF and was generally reserved for severe hospitalised patients with UC. Infliximab demonstrated efficacy as a rescue therapy in patients with acutely severe or moderately severe UC attacks not responding to conventional therapy. In a randomised double-blind placebo-controlled study, patients who fulfilled the index criteria for fulminant UC or a moderately severe acute attack of UC were randomised to receive infliximab or placebo after inadequate response to corticosteroids. Within 3 months, significantly fewer patients receiving a single dose of infliximab required colectomy compared with those receiving placebo (7/24 (29%) vs. 14/21 (67%), P = 0.017). Colectomy rates based on severity of disease were lower in patients with less severe UC for infliximab than with placebo (0% vs. 63%, P =0.009).[2] Recently, the efficacy of cyclosporin was compared with infliximab for acute severe UC for 98 days in a randomised, parallel, open-label trial. Patients (n = 115) who had an acute severe flare of UC (Lichtiger score >10) and were refractory to IV steroids were randomly assigned to IV cyclosporin (2 mg/kg per day for 1 week, followed by oral drug until day 98) or infliximab (5 mg/kg on days 0, 14, and 42). All patients with a clinical response at day 7 received AZA. The primary endpoint was treatment failure at any time. Failure was defined by the absence of a clinical response at day 7, a relapse between day 7 and day 98, absence of steroid-free remission at day 98, a severe adverse event leading to treatment interruption, colectomy or death. Treatment failure occurred in 35 (60%) patients in the cyclosporin group and 31 (54%) patients in the infliximab group (P = 0.52).[3] Although this was a recent study evaluating an anti-TNF agent for acute severe UC, the majority of anti-TNF experience can be observed in patients with moderate-to-severe UC.


In 2006, infliximab was the first anti-TNF biologic approved to treat UC. It was approved in the EU for treatment of moderately to severely active UC in patients who have had an inadequate response to conventional therapy, including corticosteroids and mercaptopurine or AZA, or are intolerant of or have medical contraindications for such therapies.[4] The source of data for the approval was based on the pivotal ACT clinical trials. ACT 1 and ACT 2 were randomised, placebo-controlled, double-blind trials that evaluated the safety and efficacy of infliximab for induction and maintenance therapy in adults with UC. Each study included 364 patients with moderate-to-severe active UC despite treatment with concurrent medications. Patients received placebo or infliximab infusions (5 or 10 mg/kg) at weeks 0, 2 and 6 and then every 8 weeks until week 46 in ACT 1 and week 22 in ACT 2. Patients were followed for 54 weeks in ACT 1 and 30 weeks in ACT 2. Also patients on concomitant thiopurines without steroids were allowed to enter the studies. Clinical response, clinical remission and mucosal healing were assessed at weeks 8 and 30 in both studies and at week 54 in ACT 1. Mucosal healing was defined as endoscopy subscore of 0 or 1.[5]

Infliximab was superior to placebo in achieving clinical response/remission, mucosal healing and having corticosteroid-sparing effects in patients. Patients who received infliximab were significantly more likely to have a sustained clinical response at weeks 8, 30 and 54 compared with placebo (ACT 1). Table 1 summarises these results. The ACT trials demonstrated the ability of patients to achieve corticosteroid-free remission. In ACT 1, 61% of patients were receiving corticosteroids at baseline (median daily dose 20 mg/day). The proportion of patients who were in clinical remission and had discontinued corticosteroids at week 30 was significantly higher with infliximab 5 mg/kg than with placebo (24.3% vs. 10.1%; P = 0.03), and this was maintained to week 54 (25.7% vs. 8.9%; P = 0.006).[5]

Table 1. ACT 1 and 2 trials: Summary of results[5]
OutcomePlacebo (n = 121)IFX 5 mg/kg (= 121)Placebo (n = 123)IFX 5 mg/kg (n = 121)
  1. IFX, infliximab.

  2. a


  3. b

     = 0.003.

  4. c

     = 0.030.

  5. d

     = 0.010.

  6. e

     = 0.006.

  7. f

     = 0.009.

  8. g

     Absolute Mayo subscore for endoscopy of 0 or 1.

Clinical response, n (%)
Week 3036 (29.8)63 (52.1)a 32 (26.0)57 (47.1)a
Week 5424 (19.8)55 (45.5)a
Clinical remission, n (%)
Week 3019 (15.7)41 (33.9)a 13 (10.6)31 (25.6)b
Week 5420 (16.5)42 (34.7)a
Steroid-free remission, n (%)
Week 308/79 (10.1)17/70 (24.3)c 2/60 (3.3)11/60 (18.3)d
Week 547/79 (8.9)18/70 (25.7)e
Mucosal healingg, n (%)
Week 3030 (24.8)61 (50.4)a 37 (30.1)56 (46.3)f
Week 5422 (18.2)55 (45.5)a

With regard to mucosal healing, the ACT trials showed that infliximab rapidly induces mucosal healing in patients with UC (Table 1). In ACT 1, infliximab-treated patients were significantly more likely than placebo patients to achieve mucosal healing (Mayo endoscopic subscore 0 or 1) at weeks 8, 30 and 54. At week 8, 62% of patients treated with infliximab 5 mg/kg achieved mucosal healing compared with 33.9% of patients in the placebo group (P < 0.001). The benefit of infliximab was sustained at week 30 (50.4% vs. 24.8% for infliximab and placebo respectively; P < 0.001) and week 54 (45.5% vs. 18.2% respectively; P < 0.001).[5] In both ACT 1 and ACT 2, the proportion of patients reporting any adverse event was similar among the three treatment groups (infliximab 5 mg/kg, infliximab 10 mg/kg and placebo). The incidence of infections was also similar among the groups in both studies. The placebo-treated group had a higher proportion of subjects with worsening UC and anaemia.[5] Of the 728 patients evaluated in ACT 1 and ACT 2, 630 (87%) patients had complete colectomy follow-up. The cumulative incidence of colectomy through 54 weeks was 10% for the combined infliximab group and 17% for placebo (P = 0.02).[6] Patients who achieved benefit with infliximab in the main ACT 1 and ACT 2 trials were eligible to participate in the long-term open-label extension studies and received up to three years of additional therapy. The extension studies assessed long-term efficacy, quality of life (QoL) and safety. The authors concluded that patients who entered the extension studies maintained clinical benefit and health-related QoL up to an additional three years of infliximab therapy and no new safety signals were observed.[7]

In addition to the controlled trial experience with infliximab, there are some data that were established from local real-life cohorts from centres in Belgium, France and Italy.

Belgium experience

One long-term infliximab experience comes from a single-centre community cohort based out of Leuven, Belgium. Ferrante et al. conducted a cohort study to determine the long-term (median follow-up 33 months) effects of infliximab therapy in patients with refractory UC and to identify predictors of sustained clinical response and colectomy. Patients (n = 121) received a single infliximab dose (5 or 10 mg/kg) or infliximab induction (infusions at weeks 0, 2 and 6). The majority of patients who achieved a clinical response then received scheduled maintenance infliximab therapy every 8 weeks (75%), whereas the remaining patients received episodic therapy. Patients with all types of extension of colitis were included.[8]

An initial short-term clinical response was achieved in 63% of patients (n = 121). Of these, 41% had a complete response (absence of diarrhoea and blood) and 22% had a partial response (marked clinical improvement but persistent rectal blood loss). From the overall cohort (n = 121), 45 patients had an absence of short-term clinical response. One patient who received the induction scheme achieved clinical response after a fourth infusion and four patients who did not receive the induction scheme achieved clinical response after a second infusion. Short-term mucosal healing was assessed in 70 patients and 53% overall, achieved short-term mucosal healing (Mayo endoscopic subscore of 0 or 1). Among the 81 patients achieving a clinical response, 68% had a sustained clinical response during a median (interquartile range) follow-up of 33.4 months (Table 2).[8]

Table 2. Long-term outcomes with infliximab from the Belgian cohort (= 121 unless otherwise stated)[8]
OutcomeProportion of patientsa
  1. a

     Approximately a median follow-up time of 33 months.

  2. b

     = 81; those who achieved clinical response at a certain time.

Sustained clinical responseb68%

French experience

Oussalah et al. conducted a retrospective study that involved five French academic centres to assess the short- and long-term efficacy and safety of infliximab in a large series of patients with UC. Between January 2000 and August 2009, 191 patients received at least one infliximab infusion for UC with a median follow-up per patient of 18 months. Approximately 18% of patients (35/191) received infliximab as a rescue therapy after cyclosporin failure. Regular treatment with infliximab was prescribed in approximately half of the patients (97/191). In 128 patients (67%), infliximab was used to treat active disease, including acute severe colitis, in patients refractory to corticosteroids, and patients who failed immunomodulators. Outcomes assessed included primary nonresponse to infliximab, infliximab optimisation, infliximab failure, colectomy and UC-related hospitalisations (Table 3).[9]

Table 3. Outcomes with infliximab from the French cohort (= 191 unless otherwise stated)[9]
OutcomeProportion of patients
Primary nonresponse (after induction therapy)22%
Need for optimisation (n = 80)45%
Probability of maintaining optimisation-free clinical benefit at:
1 year98.7% (±1.2%)
2 years46.0% (±7.0%)
3 years41.4% (±7.6%)
6 years26.8% (±8.4%)
Sustained clinical benefit (n = 149; responders to induction)55.7%
Probability of maintaining failure-free survival
1 year66.3% (±4.2%)
2 years52.6% (±4.9%)
3 years40.7% (±5.8%)
6 years27.1% (±7.7%)
Probability of maintaining colectomy-free survival
1 year83.5% (±2.96%)
2 years77.8% (±3.6%)
3 years75.7% (±4.1%)
6 years61.4% (±10.6%)
Need for UC-related hospitalisations36.1%
Probability of maintaining hospitalisation-free survival
1 year66.7% (±3.7%)
2 years60.2% (±4.1%)
3 years57.1% (±4.4%)
6 years44.6% (±8.3%)

Italian experience

Armuzzi and colleagues conducted an open, uncontrolled study in 126 patients (n = 36 retrospectively included and n = 90 prospectively included starting from 2007) to evaluate the efficacy of infliximab in steroid-dependent UC. Patients included in this study were part of three IBD Units in Italy and were treated with standard infliximab induction and maintenance regimens. Concomitant thiopurines were initiated or maintained based on clinical judgment. Co-primary prespecified endpoints included steroid-free clinical remission at 6 and 12 months, steroid-free clinical remission and mucosal healing at 12 months and colectomy within 12 months. Results for the co-primary endpoints are summarised in Table 4. The group also conducted subanalyses to determine the impact if patients were thiopurine-naïve or exposed and if patients were treated with combination therapy compared with infliximab monotherapy. Subanalyses showed that thiopurine-naïve patients had higher rates of steroid-free clinical remission at 6 months and steroid-free clinical remission with endoscopic remission at 12 months. In addition, combination therapy (infliximab + AZA) was more effective than infliximab monotherapy up to the first 6 months.[10]

Table 4. Outcomes with infliximab from the Italian cohort (n = 126)[10]
Outcome6 months12 months
Steroid-free clinical remission53.2%46.8%
Steroid-free clinical remission + mucosal healingN/A32.5%

Clinical experience with infliximab for treating UC has ranged from randomised controlled trials to long-term real-life cohorts. However, other anti-TNFs are now available or being evaluated for the treatment of UC.

Present: What does anti-TNF treatment in UC currently look like?


To add to the arsenal of treatments for UC, another anti-TNF is currently available and approved to treat UC. In April 2012, adalimumab was approved in Europe for the treatment of moderately to severely active UC in adult patients who have had an inadequate response to conventional therapy.[11] The key studies that evaluated the efficacy of adalimumab in UC are the induction and maintenance trials, ULTRA 1 and ULTRA 2 respectively.

ULTRA 1 was a randomised, controlled, double-blind, 8-week induction study with an open-label phase through week 52.[12] ULTRA 2 had a similar design but included induction followed by maintenance through week 52. ULTRA 2 was slightly different compared with ULTRA 1 as ULTRA 2 evaluated patients through week 52 but did not have an open-label phase after induction like in ULTRA 1.[13]

ULTRA 1 included three treatment arms (n = 130 in each arm): adalimumab 160/80 (160 mg at week 0, 80 mg at week 2, 40 mg at weeks 4 and 6); adalimumab 80/40 (80 mg at week 0, 40 mg at weeks 2, 4 and 6) and placebo. The primary endpoint was clinical remission (Mayo score ≤2 with no individual subscore >1) at week 8. The patient population was similar to ACT 1 and 2: moderately to severely active UC, Mayo score of 6–12 and endoscopy subscore of 2–3 and concurrent treatment with at least oral corticosteroids, mercaptopurine/AZA or failed to respond or could not tolerate previous corticosteroids or immunomodulators.[12]

The primary endpoint of clinical remission was achieved with the higher adalimumab dose of 160/80 (18.5% vs. 9.2%; = 0.031 vs. placebo). The proportion of patients in clinical remission at week 8 in the placebo and ADA 80/40 groups was similar (9.2% vs. 10%). Secondary endpoints from ULTRA 1 are summarised in Table 5. The authors commented that the high placebo rates for the secondary endpoints may have contributed to the lack of statistical significance. Adalimumab treatment was generally well tolerated at both doses and the safety profile was comparable to that of placebo. UC was the most common adverse event leading to discontinuation in 4.0% of placebo, 3.8% of adalimumab 80/40 and 3.6% of adalimumab 160/80 patients. The incidence of injection site pain was low and similar across the three study groups.[12]

Table 5. ULTRA 1 and ULTRA 2: Summary of results[12, 13]
OutcomePlacebo (n = 130)ADA 80/40 (n = 130)ADA 160/80 (n = 130)Placebo (n = 246)ADA 160/80 (n = 248)
  1. ADA, adalimumab.

  2. a

     = 0.031.

  3. b

     < 0.05.

  4. c

     < 0.005.

  5. d

     = 0.038.

  6. e

     = 0.006.

  7. f

     = 0.035.

  8. g

     = 0.028.

Clinical remission,%
Week 89.2%10.0%18.5%a 9.3%16.5%b
Week 528.5%17.3%c
Clinical response,%
Week 844.6%51.5%54.6%34.6%50.4%c
Week 5218.3%30.2%b
Mucosal healing,%
Week 841.5%37.7%46.9%31.7%41.1%b
Week 5215.4%25.0%b
Rectal bleeding subscore ≤1
Week 866.2%70.0%77.7%d 58.1%70.2%e
PGA subscore ≤1
Week 846.9%53.8%60.0%f 37.4%46.0%
Stool frequency subscore ≤ 1
Week 837.7%36.2%48.5%28.5%37.9%g

ULTRA 2 included two treatment arms: adalimumab 160/80 (n = 248; 160 mg at week 0, 80 mg at week 2, and 40 mg every other week starting at week 4) and placebo (n = 246). Patients were stratified based on prior exposure to anti-TNF treatment and randomly assigned to treatment groups. The co-primary endpoints were remission at week 8 and 52. The patient population was similar to ULTRA 1.[13]

Overall, clinical remission at week 8 was achieved in 16.5% of patients on adalimumab compared to 9.3% on placebo (P = 0.019); values for week 52 for the adalimumab and placebo groups were 17.3% and 8.5% respectively (P = 0.004). When stratifying the results for patients who were naïve to anti-TNF treatment, the values in each group were slightly higher. The rates of remission at week 8 were 21.3% on adalimumab and 11% on placebo (P = 0.017); values for week 52 for the adalimumab and placebo groups were 22% and 12.4% respectively (P = 0.029). The values for each treatment group were lower when analysing those who had previously received anti-TNF agents. Rates of remission at week 8 were 9.2% on adalimumab and 6.9% on placebo (P = 0.559); corresponding values for week 52 were 10.2% and 3% (P = 0.039). Secondary endpoints included clinical response and mucosal healing and the differences between adalimumab and placebo were significant at weeks 8 and 54 in favour of adalimumab (Table 5). The results also favoured adalimumab for the secondary endpoints in patients who were naïve to anti-TNF treatment. The differences between adalimumab and placebo did not reach significance for the majority of secondary endpoints in patients previously exposed to anti-TNF treatment.[13] A subgroup analysis of ULTRA 2 assessed the one-year maintenance outcomes in patients who responded to induction therapy with adalimumab. Patients who achieved clinical response at week 8 were assessed at week 52 to determine if they achieved several outcomes such as clinical remission, clinical response and mucosal healing. At week 8, approximately half of the adalimumab-treated patients (123/248, 49.6%) achieved clinical response. Of those, 30.9%, 49.6% and 43.1% achieved clinical remission, clinical response and mucosal healing, respectively, at week 52. In addition, of the week 8 responders who were using corticosteroids at baseline (n = 90), 21.1% achieved steroid-free remission and 37.8% had discontinued steroids at week 52.[14]

One of the main differences in the induction results between ULTRA 1 and ULTRA 2 was the high placebo response rate observed in ULTRA 1, whereas the corresponding rates in ULTRA 2 were similar to those observed in ACT 1 and 2. The authors commented that adalimumab 160/80 is an effective induction dose with 40 mg every other week effective as maintenance treatment. It is unknown if higher induction doses will result in improved efficacy. In addition, it is also unknown if a higher maintenance dose in ULTRA 2, such as 40 mg weekly, would have resulted in improved efficacy. A Crohn's disease dose-finding maintenance trial did not show any difference between 40 mg every other week compared with weekly administration.[13]

Apart from the ULTRA trials, there is additional clinical experience on the use of adalimumab for UC. Data from a study on a small cohort of patients (n = 30) from Spain demonstrated that adalimumab induction and maintenance therapy was effective in patients who previously failed other therapies including infliximab, At weeks 4 and 12, clinical response was achieved in 16 (53%) and 18 (60%) of patients, respectively, and clinical remission was achieved in three (10%) and eight (27%) of patients respectively. Adalimumab was continued in 15 (50%) of patients after 48 weeks (mean) follow-up. Colectomy was required in six (20%) of patients; however, patients who achieved clinical response at week 12 avoided colectomy over the long term.[15] Real-life data on the use of anti-TNF agents in UC was obtained from single-centre Canadian cohort. This was a prospective study with a long-term follow-up of 53 patients treated with either infliximab or adalimumab. Efficacy was assessed using physician's global assessment focusing on normalisation of bowel frequency, absence of blood with defecation and tapering of corticosteroids until discontinuation. Responses to induction therapy were 96.4% (27/28) for infliximab and 80% (20/25) for adalimumab (P = N.S.). Responses to maintenance therapy were similar: infliximab 77.8% (14/18) and adalimumab 70.0% (14/20; P = N.S.).[16]

Overcoming misunderstandings in the management of UC

With the current data and clinical experience available in UC, especially with anti-TNF therapy, clinicians have been able to clarify many of the misunderstandings around the management of UC. Many of these misunderstandings have been outlined in the review by Ochsenkuhn and D'Haens.[17] Relating to this review, one of the main misunderstandings is that anti-TNF therapy is less effective in UC compared with CD. Using the example of infliximab since it has the most abundant data and clinical experience, findings from the pivotal CD trial ACCENT I[18] demonstrated similar rates of clinical remission and mucosal healing compared with the ACT 1 and 2 for UC.[5] The other misunderstanding is that anti-TNF therapy should be reserved as rescue therapy in acute severe UC. Although there is experience with demonstrated efficacy of infliximab in this severe UC population, the vast majority of clinical data for both infliximab and adalimumab come from patients who were nonhospitalised with moderate-to-severe UC.[5, 8, 12, 13]

Applying experience with anti-TNFs in CD to UC

Infliximab and adalimumab were approved to treat CD prior to UC approval which results in a more pronounced clinical experience with these agents in the former. Thus, clinicians might be inclined to extrapolate the knowledge with anti-TNFs from CD to UC. Taking the example of infliximab; the SONIC trial was unique as it was the first study in CD to evaluate the efficacy of an anti-TNF treatment strategy compared with AZA. Briefly, SONIC was a multicentre, randomised, double-blind, 30-week, active-controlled trial with a 20-week blinded extension that compared the efficacy of AZA monotherapy, infliximab monotherapy, and infliximab plus AZA combination therapy in patients with active Crohn's disease who were naïve to immunomodulators and anti-TNF agents. Therefore, patients received anti-TNF therapy earlier in their course of disease after 5-ASA and steroids. Patients also had moderate-to-severe active CD and were considered to have early disease (median disease duration 2.2 years). The study found that an infliximab-based treatment strategy (infliximab alone or in combination with AZA) was superior to AZA monotherapy.[19]

How does the experience from SONIC apply to UC? Preliminary data are available from a UC study that had a similar design to SONIC. UC SUCCESS included 239 patients with moderate-to-severe UC who were naïve to biologics, were failing corticosteroids and were either naïve to AZA or had stopped AZA ≥3 months before entry. Patients were randomised to receive AZA 2.5 mg/kg, infliximab 5 mg/kg or infliximab 5 mg/kg plus AZA 2.5 mg/kg for 16 weeks. The primary endpoint was steroid-free remission at week 16 and secondary endpoints included response and mucosal healing both at week 16. Preliminary results showed that the primary endpoint was achieved in the infliximab plus AZA arm compared with the AZA arm. In addition, infliximab plus AZA was superior to AZA or infliximab monotherapy in inducing steroid-free remission in patients (Table 6). Patients treated with an infliximab-based strategy were more likely to achieve response and mucosal healing than those treated with AZA monotherapy.[20] What does UC SUCCESS mean for anti-TNF therapy moving forward? It is difficult to make any specific conclusions and more data are needed, however, the limited data (only from infliximab experience) point that anti-TNF therapy in combination may be more effective in early UC compared with monotherapy at 16 weeks.

Table 6. UC SUCCESS: Preliminary results[20]
OutcomesInfliximab + AZA (n = 78)Infliximab (n = 77)AZA (n = 76)
  1. AZA, azathioprine.

  2. a

     < 0.05 compared to infliximab.

  3. b

     < 0.05 compared to AZA.

Clinical endpoints, week 16 (proportion of patients)
Mucosal healing63%b55%b37%

Predictors of good treatment outcome

Currently abundant clinical data are available on the two anti-TNF treatments for UC and both infliximab and adalimumab are effective. No head to head studies have been conducted to indicate which anti-TNF is more effective and the choice at times is determined by physician and patient preference. Anti-TNF treatment is not for everyone but certain patients may benefit more than others. Having predictors of good treatment outcome will help identify those who would benefit most from anti-TNF treatment. Indicators for poor disease prognosis will also help stratify those who may benefit from earlier treatment with an anti-TNF treatment.

There are limited data around predictors of good treatment outcome in UC. Results from the real-life cohort of patients treated with infliximab from Leuven mentioned previously, in a univariate analysis, predictors of sustained clinical response were a significant short-term C-reactive protein drop and extensive colitis at baseline. Patients with concomitant use of immunosuppressive agents demonstrated a trend towards a better sustained clinical response.[8] Similar findings were observed in the Italian cohort demonstrating that factors associated with steroid-free clinical remission at 6 and 12 months included thiopurine-naïve status, combination therapy and a drop in serum CRP levels to normal value.[10] Another UC study showed moderately severe, nonhospitalised patients achieve better sustained results (response and remission) on infliximab compared with those with acute severe disease. The same study by Seow et al., assessed clinical outcomes according to detectable or undetectable trough serum infliximab levels in UC patients on maintenance infliximab. The authors concluded that patients treated with infliximab, a detectable trough serum infliximab predicts clinical remission, endoscopic improvement and a lower risk for colectomy.[21] The London Position Statement on Biological Therapy in IBD provides some commentary on predictors of response. They suggest that identifying factors that influence response or failure could help select the most appropriate biological regimen for individual patients.[22] A few studies have examined predictors of (non) response to anti-TNF therapy in UC and factors associated with short-term response include young age and absence of antibodies.[22] There is also some data on the use of genes to predict response to infliximab in UC, Arijs and colleagues assessed the predictive response signature for infliximab using colonic mucosal gene expression. Pre-treatment colonic mucosal expression profiles were compared for responders and nonresponders that determined the predictive panel of genes.[23]

Predictors for poor disease prognosis

The natural course of UC progression for a given patient is difficult to predict. Some patients can experience a course of disease that can lead to poor outcomes such as hospitalisations, colectomy and death. Still questions remaining on how to identify the right patient for earlier treatment with anti-TNF treatment, however, increasing experience is helping to change the paradigm. The majority of anti-TNF clinical experience in UC is with infliximab and therefore there are some data from different sources (local cohort data from Leuven, Belgium and France) on the predictive factors of colectomy.[8, 9] Results from the Leuven single-centre community cohort study mentioned previously, identified the following as independent predictors of need for colectomy: absence of short-term clinical response (Hazard ratio [HR] 10.8 (95% confidence interval [95% CI] 3.5–32.8), < 0.001], baseline CRP level ≥5 mg/L [HR 14.5 (95% CI 2.0–108.6), P = 0.006] and previous IV treatment with corticosteroids and/or cyclosporin [HR 2.4 (95% CI 1.1–5.9), P = 0.033].[8] Results from the French experience also described earlier, provided some data on predictors of colectomy (from a multivariate analysis without haemoglobin as covariate): no clinical response after infliximab induction (HR 7.06 [95% CI 3.36–14.83], < 0.0001), CRP >10 mg/L after infliximab induction (HR 5.11 [95% CI 1.77–14.76], P = 0.003), and previous treatment with cyclosporin (HR 2.53 [95% CI 1.22–5.28], P = 0.01).[9]

These predictive factors may help provide some guidance to clinicians as a threshold to use infliximab therapy earlier. More data are still needed to help identify patients who will benefit the most form anti-TNF treatment and those who will benefit from earlier treatment. Hopefully the near future will hold some of these solutions that will help move the management of UC forward.

What does the future hold for anti-TNF agents in UC?

In a short period of time, since 2006, anti-TNF therapy in UC has come a long way. Still some challenges exist such as recognition that up to now, UC has not been taken as a serious disease. Ochsenkuhn and D'Haens mention in their review of misunderstanding in UC that UC is still considered a more ‘benign’ disease than Crohn's disease.[17] This ‘benign’ assumption may lead to multiple rounds of ineffective therapy which can delay anti-TNF therapy to a point where it could be too late and the efficacy in patients be reduced. The authors go on to say that ‘the drugs that we use and the way in which we use them do not bring the majority of affected patients into ‘remission’, which is the most desirable outcome.’

The results from the infliximab and adalimumab clinical trial experience demonstrate efficacy based on specific endpoints such as remission, response and mucosal healing. Some of the trials also reported impact on quality of life. However, little or no data exist on anti-TNF therapy and the impact on disability. Functioning and disability are becoming more important in assessing the impact of chronic diseases and the effectiveness of treatment on the natural course of the disease. The first IBD disability index has been developed and will be useful for clinical practice, disease modification trials and health reporting in IBD.[24]

Identifying patients with poor prognosis may allow for earlier intervention. As mentioned previously, a few studies have identified predictors for colectomy which may warrant earlier treatment with infliximab. More investigation is needed, but preliminary research shows the possibility of a risk matrix model for the prediction of colectomy.[25] A risk matrix tool can provide a quick guide for clinicians on when and in whom to start anti-TNF therapy. Advancing through the UC-treatment algorithm more rapidly using time-structured reassessment at each stage of therapy (accelerated step-up) may help ensure earlier effective anti-TNF therapy and that therapeutic goals are achieved.

Over time, the treatment goals have evolved from being symptom free to long-term remission and reduction of steroid use. The importance of mucosal healing is better understood and more data are coming out on the association of mucosal healing and improved outcomes. Effective therapies, like anti-TNFs, have helped change the perception of what can be achieved for patients, but what does the future hold as the ultimate aim for treatment-treat to heal? Is it the ultimate goal and are clinicians confident that the current anti-TNF agents will allow patients to achieve it?

The future will help change the landscape of anti-TNFs. Another anti-TNF currently under development for UC is golimumab. Golimumab is already approved in the US and Europe to treat certain rheumatologic indications.[26] Preliminary results from the PURSUIT induction trial (PURSUIT SC) demonstrated efficacy of golimumab in UC. The patient population of PURSUIT SC was similar to ACT 1 and 2 and patients were naïve to anti-TNF therapy. PURSUIT SC had an adaptive design with Phase 2 dose ranging followed by a confirmatory Phase 3 component. Patients randomised after dose selection (n = 774) were treated with either placebo, golimumab 200 mg/100 mg or golimumab 400 mg/200 mg at weeks 0 and 2. The primary endpoint was clinical response at week 6 in patients enrolled after dose selection. Secondary endpoints included clinical remission and mucosal healing. The PURSUIT SC induction results showed significant differences in efficacy in favour of both golimumab regimens compared with placebo (Table 7).[27] The initial data on golimumab in UC look promising and may be added to the approved list of anti-TNF therapies for UC in the near future.

Table 7. PURSUIT SC: Summary of results at week 6 (among patients randomised after dose selection)[27]
OutcomePlacebo (n = 256)200 mg/100 mg (n = 257)400 mg/200 mg (n = 258)
  1. a

     < 0.0001.

  2. b

     = 0.0005.

Clinical response29.7%51.8%a55.0%a
Clinical remission6.3%18.7%a17.8%a
Mucosal healing28.5%43.2%b45.3%a

In a prospective, open-label study of moderately to severely active UC receiving certolizumab pegol.

400 mg at weeks 0, 2, 4, as an induction regimen followed by 400 mg every 4 weeks as a maintenance regimen, 7 of 10 patients achieved clinical response, as measured by the total Mayo score, while complete clinical remission was observed in only one patient at week 14.[28] These results are difficult to interpret due to the lack of control arm. Large randomised controlled trials are awaited to assess efficacy and safety of certolizumab pegol in UC. In clinical practice, treatment choice should be guided by physician and centre experience.


In a short period of time, the availability and use of anti-TNFs have helped change the therapeutic approach to UC (Figure 1). The use of anti-TNFs in UC has demonstrated efficacy in achieving steroid-free remission and mucosal healing, and in changing the natural history (colectomies). Should we be moving beyond these goals and try to achieve new endpoints such as improvement in function or disability or ultimately treating to heal? More data are needed to help answer these clinical questions but the field is heading in the right direction with the development of a disability index. Although the management of UC with anti-TNFs has come a long way, still CD appears to be ahead. In CD, it has become an accepted strategy to treat patients early and aggressively, which can lead to improved outcomes. We propose here a new treatment algorithm in patients with refractory UC (Figure 2). With UC being a progressive disease with long-term structural and functional complications, can the concept of early aggressive management apply? This has yet to be determined but this concept may lead to studies where anti-TNFs are evaluated in early UC.

Figure 1.

Evolution of anti-TNFs in UC.

Figure 2.

A proposed algorithm for the treatment of ulcerative colitis.


Guarantor of the article: S. Danese.

Author contributions: Silvio Danese wrote the first draft that was shared with and edited by WR. JFC, LPB and PR. All authors approved the final version of the manuscript.


Declaration of personal interests: Silvio Danese has served as a speaker, a consultant and an advisory board member for Schering-Plough, Abbott Laboratories, Merck & Co, UCB Pharma, Ferring, Cellerix, Millenium Takeda, Nycomed, Pharmacosmos, Actelion, Alphawasserman, Genentech, Pfizer, Astra Zeneca, Novo Nordisk, Cosmo Pharmaceuticals, Vifor, and Johnson and Johnson.

Laurent Peyrin-Biroulet, received consulting and lecture fees from Abbott and Merck and consulting fees from UCB-pharma.

Jean-Frédéric Colombel has served as a consultant and an advisory board member for Abbott Laboratories, Abbott Park, Illinois; Amgen, Thousand Oaks, CA; Biogen Idec Inc, Cambridge, M; Boehringer-Ingelheim, Inc., Ridgefield, CT; Bristol Meyers Squibb, Princeton, NJ; Cellerix SL, Madrid, Spain; Chemocentryx, Inc.Mountain View, CA; Centocor, Malvern, PA; Cosmo Technologies, Ltd, Milano, Italy; Elan Pharmaceuticals, Inc., San Francisco, CA; Genentech, San Francisco, CA; Giuliani SPA, Milano, It; Given Imaging, Hamburg, Ge; Glaxo Smith Kline, Research Triangle Park, NC; Immune Pharmaceuticals Ltd., ISR; Merck & Co., Inc., NJ, USA; Millenium Pharmaceuticals Inc., Cambridge, MA; Neovacs SA, Paris, F; Ocerra Therapeutics, Inc. (previously named Renovia, Inc.), San Diego, CA; Pfizer Inc. New York, USA; Prometheus, San Diego, CA; Sanofi, USA; Schering Plough Corporation, Kenilworth, NJ; Shire Pharmaceuticals, Wayne, PA; Synta Pharmaceutical Corporation, Lexington, MA; Takeda, Philadelphia, PE; Teva Pharmaceuticals, North Wales, PA and Petah Tikva, Israel; Therakos, Exton, PA; TXcell, Valbonne, Fra; UCB Pharma (previously named Celltech Therapeutics, Ltd.), Atlanta GA and Brussels, Belgium; Wyeth Pharmaceuticals, Collegeville, Pen.

Walter Reinish has been consultant for Aesca, Abbott, Biogen Idec, Centocor, Ferring, Genentech, Millennium Research Group, MSD, Novartis, Schering-Plough, Shire Pharmaceuticals, and UCB Pharma; advisory committees for Aesca, Abbott, Biogen Idec, Centocor, Genentech, Millennium Research Group, MSD, Novartis, Schering-Plough, Shire Pharmaceuticals, and UCB; lecture fees from Ferring.

Paul Rutgeerts has served as a speaker for Centocor, Schering-Plough, UCB, Abbott, Elan-Biogen, a consultant for Centocor, Schering-Plough, UCB, Abbott, Elan-Biogen, NovImmune, Italfarmako, Bristol-Myers Squibb, Millennium Pharmaceuticals, Tillots, GSK and ChemoCentryx and has received research funding from Centocor, Schering-Plough, UCB and Abbott.

Declaration of funding interests: None.