Background: Patient care in ulcerative colitis (UC) remains challenging despite an array of established treatment options and emerging new therapies. The management of UC therapy should be guided by the endoscopic extent of inflammation, disease severity, and prognostic factors of poor outcome. Complete remission, defined as durable symptomatic and endoscopic remission without corticosteroid therapy, is the desired treatment goal. Summary: This review focuses on treatment recommendations for different clinical scenarios in moderate-to-severe UC: Active UC of any extent not responding to aminosalicylates, steroid-dependent UC, steroid-refractory UC, immunomodulator-refractory UC, and acute severe UC. Comprehensive treatment algorithms for daily clinical practice were developed based on published guidelines and current literature. Key Messages: While current treatment options including a number of biologicals and small molecules have evolved UC treatment to achieve sustained remission in a majority of patients, upcoming treatment options with different molecular pathways and different modes of actions will further increase the need for personalized medicine.
Ulcerative colitis (UC) is characterized by a chronic immune-mediated inflammation of the colon. The cause of the aberrant immune response remains largely unknown, but dietary and environmental risk factors have a role as well as host factors such as genetic susceptibility and gut microbiota . The annual incidence is higher in industrialized countries and has steadily increased over the last decades worldwide .
Clinical presentation of UC is characterized by a gradual or sudden onset of hematochezia, diarrhea, and abdominal pain . Symptoms may also include urgency and fecal incontinence, while weight loss and fever are typical features of severe disease activity. The presence of anemia, hypoalbuminemia, and elevated C-reactive protein (CRP) may suggest inflammatory bowel disease, elevated fecal calprotectin is a sensitive (but not specific) indicator of intestinal inflammation . The diagnosis of UC is based on a combination of clinical, laboratory, imaging, and endoscopic parameters, including histopathology. Endoscopic findings include continuous colonic inflammation characterized by erythema, loss of normal vascular pattern, granularity, erosions, friability, bleeding, and ulcerations [5, 6].
At diagnosis, most patients have left-sided UC (40%) with mild-to-moderate disease activity . The natural course of UC follows a gradual progression of disease extension which is seen in 10–30% of patients. Most patients have a chronic remitting and relapsing disease course with a 10-year cumulative risk of relapse of 70–80% [8, 9]. The 5- and 10-year cumulative risk of colectomy is 10–15% [10, 11].
Therapeutic management in UC should be guided by the endoscopic extent of inflammation, disease severity, and prognostic factors of poor outcome, for example, age <40 years at diagnosis, extensive disease, or the presence of extraintestinal manifestations and response to previous medication [12-14]. The goal of treatment is complete remission defined as durable symptomatic and endoscopic remission without corticosteroid therapy. Current data do not yet support histologic healing as a treatment goal in UC [15, 16].
The extent of UC may vary considerably and is described by the Montreal Classification . The endoscopic extent is categorized as distal to the rectosigmoid junction (E1, proctitis), extending anywhere from the sigmoid to the splenic flexure (E2, left-sided colitis or distal colitis) or extending beyond the splenic flexure (E3, extensive colitis or pancolitis). The extent of inflammation has implications on patient medical management and is also a risk factor for the development of colorectal cancer in case of left sided or extensive disease. Proximal extension of proctitis may occur in 20–50% of adult patients with UC . Disease severity also influences treatment modalities and different clinical indices of UC disease activity have been suggested (Table 1) . Most national and international guidelines recommend classifying UC into mild, moderate, and severe disease activity based on the Truelove and Witts’ Severity Index [12, 14, 20, 21]. There is no fully validated definition of remission, but it is suggested to use a combination of clinical parameters (stool frequency ≤3/day with no bleeding) and no mucosal lesions at endoscopy .
Active UC of Any Extent Not Responding to Aminosalicylates
In mild-to-moderate UC of any extent, aminosalicylates (5-ASA) such as mesalamine are the preferred initial treatment [12-14] (Fig. 1). In patients with disease activity limited to the rectum, topical therapy alone might suffice, but combination therapy is more effective and is also recommended for left-sided and extensive UC [22, 23]. Table 2 gives an overview on current treatment options.
Except in case of isolated proctitis, where topical corticosteroids alone may be considered, treatment with oral corticosteroids should be initiated in patients who do not respond adequately to 5-ASA . In case of isolated proctitis, topical corticosteroids alone might be considered. The introduction of corticosteroids should be a shared decision-making process that includes patient’s preference of therapy and tolerance to 5-ASA. It is, however, recommended to start corticosteroids in patients with sustained rectal bleeding for 2 weeks, persistent abdominal symptoms after 6 weeks of adequate therapy with 5-ASA or if symptoms deteriorate . In selected cases, a prolonged therapy with up to 16 weeks might still be able to achieve remission.
Both conventional oral corticosteroids and budesonide have shown to be superior to placebo and 5-ASA in inducing remission in active UC [24-26]. Budesonide is a second-generation corticosteroid with an ileo-colonic release mechanism and low systemic bioavailability. Recently, a new formulation of budesonide multimatrix (MMX) with a colonic delivery technology to permit the release of budesonide at a controlled rate throughout the colon has been introduced. Budesonide-MMX is recommended in patients with left sided or extensive UC with mild-to-moderate disease activity who are intolerant or refractory to 5-ASA [13, 27]. It improved clinical and endoscopic parameters more likely than placebo but direct comparison in efficacy against conventional corticosteroids are lacking [28, 29]. Also, budesonide-MMX seems to be more efficacious in left-sided colitis, and oral non-MMX budesonide is not superior to placebo in the treatment of UC . However, topical non-MMX budesonide is effective in inducing remission in UC confined to the rectum [31, 32]. As almost 50% of UC patients experience corticosteroid-related adverse events (acne, sleep- and mood disturbance, glucose intolerance, and dyspepsia), budesonide-MMX with its low systemic bioavailability should always be considered as first-line therapy in patients not responding to 5-aminosalicylates (5-ASA).
In patients not responding to budesonide-MMX or with moderate-to-severe UC disease activity, conventional oral corticosteroid therapy (0.75–1 mg/kg oral prednisone-equivalent, not >60 mg/day) should be considered [33, 34]. After 2 weeks, patients should be assessed, and the daily dosage lowered by 5–10 mg every week in case of clinical response. A course of 8 weeks therapy is appropriate as to prevent early relapse.
In patients not responding adequately to conventional oral corticosteroids, stool should be retested for bacteria, parasites, and Clostridium difficile toxin as well as sigmoidoscopy with biopsies to rule out cytomegalovirus (CMV) should be performed . If infectious disease is excluded a course of intravenous corticosteroids may be initiated and if unsuccessful, the patient is considered to have steroid-refractory UC (see the corresponding section below). If a patient fails to taper corticosteroids below the equivalent of prednisolone 10 mg/day within 3 months or relapses within 3 months, he is considered steroid dependent (see the corresponding section below). In patients having successfully tapered corticosteroids, an attempt should be made to reestablish maintenance therapy with topic and/or oral 5-ASA [36, 37]. In case of relapse, combination therapy with corticosteroids and azathioprine (AZA)/6-mercaptopurine (6-MP) or biologic therapy should be considered [38-41].
Steroid-dependent UC defines a disease course that initially responds to oral corticosteroids but fails to taper below the equivalent of prednisolone 10 mg/day within 3 months or that relapses within 3 months after steroid discontinuation  (Fig. 2). Corticosteroid use in UC is associated with higher risk for relapse and colectomy . Given the adverse short- and long-term effects of corticosteroids, to achieve and maintain a corticosteroid-free remission remains paramount. After initial exposure to corticosteroids around two-third will require reintroduction of corticosteroids, and one-third will become steroid-dependent over time . Patients undergoing oral corticosteroid induction therapy should therefore be clinically assessed within 2 weeks in order to identify nonresponders early and to avoid delayed therapy escalation . In case of partial response, induction therapy with full-dose corticosteroids may be extended for another 2 weeks in selected patients, but if no clinical response is seen, therapy should be modified.
Prior to therapy escalation, diagnostic work-up should include stool test for infectious colitis and sigmoidoscopy to rule out CMV. Also, adherence to 5-ASA medication is a factor of relapsing disease activity and has to be discussed with the patient. The prevalence of nonadherence to medication is generally high (40–50%) and may increase to 68% in patients on >4 prescription medications [49, 50]. Other predictors of low adherence are age <40 years, alcohol consumption, and current smoking habits . Patients with once daily 5-ASA medication have higher remission rates than t.i.d. (86 vs. 73%, p = 0.03), and strategies to optimize therapy adherence should include simple dosing options .
Patients with steroid-dependent disease should be treated with an either AZA/6-MP, a tumor necrosis factor (TNF)-inhibitor (preferably combined with AZA/6-MP), or vedolizumab. The role of methotrexate within UC treatment algorithms is controversial. ECCO still lists methotrexate as therapeutic option , whereas others do not . A recent Cochrane Review summarized that there is no evidence supporting the use of methotrexate for maintenance of remission in UC  and since then, 2 additional and complementing trials have also suggested a lack of efficacy for methotrexate [54, 55].
Evidence of several studies supports the benefit of AZA for maintenance of remission in patients with UC . AZA has been shown to be more effective than 5-ASA in achieving clinical and endoscopic remission in corticosteroid-dependent UC . However, thiopurine therapy carries an increased risk for lymphoma and nonmelanoma skin cancer as well as bone marrow suppression, pancreatitis, and hepatotoxicity [57-59]. Therefore, long-term treatment with thiopurines should balance the individual patient’s response with the risk and efficacy of therapy. Before treatment with AZA/6-MP is initiated, full blood count including lymphocyte numbers must be obtained. Measurement of thiopurine-S-methyltransferase activity may be considered to guide initial dosing of AZA/6-MP and identify patients at risk for severe bone marrow suppression under therapy [60-62]. However, regular WBC counts are mandatory, as bone marrow suppression may still occur irrespective of TMPT activity [63, 64]. Therapeutic drug monitoring is performed with measurement of 6-thioguanin, a thiopurine metabolite, with higher levels increasing the risk for leucopenia and possibly aplasia. Suggested through levels are 230–450 pmol/8 × 108 red blood cells for monotherapy with AZA/6-MP and >125 pmol/8 × 108 red blood cells for combination therapy with a TNF-inhibitor . Treatment response should be evaluated as early as 12 weeks when AZA/6-MP therapy is expected to be fully effective . In summary, AZA/6-MP maintenance therapy is best suited for patients with low risk of progression who responded well to a first course of corticosteroid therapy.
In all other patients with a corticosteroid-dependent course of UC, biologic therapy with either an anti-TNF drug, alone or in combination with AZA/6-MP, vedolizumab, or tofacitinib are the preferred choices. Several studies have shown anti-TNF therapy (infliximab, adalimumab, golimumab) to be effective in achieving corticosteroid-free remission (in about 30%) compared to placebo in patients with moderate-to-severe UC [67-70]. When introducing anti-TNF therapy, combination therapy with AZA/6-MP should always be considered. In the UC SUCCESS trial in biologic-naïve patients, corticosteroid-free remission was achieved in 40% with combination therapy, compared to 22% with infliximab alone . Combination therapy with adalimumab or golimumab, also in AZA/6-MP pre-exposed corticosteroid-dependent patients, has not been studied in randomized controlled trials.
Vedolizumab is a humanized monoclonal antibody that binds to α4β7 integrin on gut-homing lymphocytes and therefore inhibiting endothelial cell adhesion and migration into the intestinal wall. It has been shown to be effective to induce and maintain corticosteroid-free remission in both anti-TNF-experienced and anti-TNF-naïve patients as well as patients with prior AZA//6-MP exposure . The efficacy in TNF-naïve patients seems to be higher and is comparable with that of anti-TNF biologics in this population . The gut-selective properties of vedolizumab with fewer systemic side effects might be advantageous for patients requiring long-term biologic therapy.
Another treatment option is Tofacitinib, an oral janus kinase inhibitor that has recently been licensed for the use in moderate-to-severe UC. Tofacitinib has been shown to induce and maintain remission after treatment failure with either oral corticosteroids, AZA, or anti-TNF therapy . Tofacitinib showed a rapid onset of action with significant improvements of symptoms (stool frequency, rectal bleeding) within 3 days . Endoscopic remission was achieved in 17–19% (compared to 4–8% with placebo) after 8 weeks.
Patients not responding to corticosteroid therapy with 0.75–1 mg/kg body weight of oral prednisolone equivalent within 4 weeks or intravenous corticosteroids for at least 1 week are defined to have steroid-refractory UC after infectious complications (coexistent CMV, C. difficile-associated disease) have been excluded . Most patients will not require hospital admission for intravenous corticosteroids. In active steroid-refractory UC, anti-TNF therapy, or vedolizumab is the primary choice as both have shown to achieve corticosteroid-free remission in UC patients receiving corticosteroids [67-70, 72]. If anti-TNF therapy is started, combination with AZA/6-MP is a therapeutic option . Similar to steroid-dependent UC, tofacitinib will be an option, especially in patients with treatment failure to immunomodulatory treatment . Alternatively, tacrolimus may be initiated. Tacrolimus has shown short-term efficacy with clinical response rates reported from 38 to 68% depending on targeted through levels  but has also been effective in reducing colectomy rate . The clinical efficacy and safety in moderate-to-severe UC are similar to anti-TNF therapy, but regular monitoring for adverse effects, including nephrotoxicity and opportunistic infections, is warranted. In patients who respond to tacrolimus, AZA/6-MP is recommended for long-term maintenance therapy , but tacrolimus might still be an option for patients who are intolerant to infliximab or AZA .
Patients having active disease or relapse despite the use of AZA (2–2.5 [max. 3] mg/kg bw/day) or 6-MP (1–1.5 mg/kg bw/day) for at least 3 months and optimized dosing regimen according to 6-thioguanin levels are considered to have immunomodulator-refractory UC . Biologic therapy with either an anti-TNF agent or vedolizumab should be considered (Fig. 3). The clinical efficacy of these compounds has been demonstrated in a number of randomized controlled trials, but comparative trials are mostly lacking to guide the choice of a particular treatment regime [67-70]. The efficacy of infliximab has been compared to other anti-TNF agents using network meta-analysis technology and has shown a (nonsignificant) trend for higher remission rates for infliximab [68, 69]. Results from the VARSITY trial have been presented recently . In patients with moderate-to-severe UC, 31.3% (n = 120/383) of patients who received IV vedolizumab achieved the trial’s primary endpoint of clinical remission, defined as a total Mayo score ≤2 with no subscore >1, at week 52, compared with 22.5% (n = 87/386) of those treated with subcutaneous adalimumab at week 52 (p = 0.0061).
Data on anti-TNFs (mainly infliximab) and AZA/6-MP combination therapy derive from one study that included UC patients that were mostly naïve to AZA/6-MP and from indirect study data [71, 79, 81]. In AZA/6-MP-experienced patients, combination therapy might reduce immunogenicity (e.g., production of antidrug antibodies, increase through levels) and increases treatment efficacy when staring anti-TNF therapy. This might reduce the risk for primary or secondary loss or response. Only limited data are available on combination therapy with vedolizumab and AZA/6-MP. Some evidence points to lower levels of antidrug antibodies with combination therapy, the implications on clinical efficacy and safety remain unclear . Combination therapy with tofacitinib is not recommended as it carries an important risk for opportunistic infections .
In case of anti-TNF failure, reinduction, or dose intensification may be an option after careful evaluation including therapeutic drug monitoring with measurement of through levels and antidrug antibodies. Switching to a different anti-TNF therapy is indicated in patients with positive antidrug antibodies, whereas switching to vedolizumab is the preferred choice in case of primary (and secondary) failure [12-14]. Data on anti-TNF therapy after primary failure of vedolizumab are currently not available, but therapy with an anti-TNF drug should still be initiated in this clinical scenario. Again, tofacitinib might be a suitable alternative.
Acute Severe UC
Approximately 20% of patients with UC experience at least one severe exacerbation during the course of their disease. Acute severe UC is diagnosed according to Truelove and Witts’ criteria (Table 1) . The definition is based on ≥6 blood stained stools daily, with 1 or more of the 4 additional criteria: hemoglobin <105 g/L, ESR >30 mm/h or CRP >30 mg/L, fever >37.8°C, and tachycardia >90 b/min. These patients must be admitted to a hospital, preferably one that provides specialized surgical backup, in case a colectomy is warranted [12-14]. Acute severe UC carries a significant morbidity and mortality (1%) .
Intravenous corticosteroids with either methylprednisolone or hydrocortisone remain first-line treatment and approximately 65% will respond  (Fig. 4). In responding patients, the choice of long-term treatment should take patients disease characteristics and comorbidities into account. It may include the use of 5-ASA, but usually consists of immunosuppression with AZA. Alternatively, biologics or small molecules may be considered.
If patients are not responding to IV corticosteroids within 3–5 days, having sustained fever, bloody diarrhea ≥4 times daily, or elevated CRP, rescue therapy with ciclosporin or infliximab must be considered [85, 86]. Only case reports exist for the use of tofacitinib in this clinical scenario . Extending IV corticosteroid therapy beyond 7–10 days has no benefit . The management of acute severe UC must be medico-surgical and requires careful, ongoing clinical evaluation of patients as delayed surgery can increase postoperative complications and mortality after 7 days [88, 89]. However, second-line medical therapy is not associated with higher mortality . Intravenous cyclosporine may be initiated, and trough levels have to be checked on day 3 [90, 91]. Hypocholesterolemia and poor renal function should be excluded prior to starting treatment, and prophylaxis against Pneumocystis jiroveci may be installed [92, 93] as a combined immunosuppression entails an increased risk of infection. Short-term response rate to ciclosporin therapy has been reported between 64 and 82% and colectomy rates range from 26 to 47% . In patients who respond to cyclosporine, overlapping AZA/6-MP maintenance therapy should be started before discharge and oral cyclosporine continued for at least 3 month as bridging therapy . In patients with acute severe UC while on therapy with AZA/6-MP respond poorly to treatment with cyclosporine [96, 97] and additionally immunosuppressants are not an option for long-term treatment after discontinuation of cyclosporine.
As an alternative to rescue therapy with cyclosporine, infliximab may be started [98-102]. However, conventional standard dosing schedule for induction therapy may not be ideal, and more intensive dosing strategies have been suggested to adequately reduce colectomy rates in acute severe UC . Intestinal protein loss with rapid clearing of infliximab but also high TNF burden, anti-TNF neutralization, and reduced tissue penetration have all been proposed to attribute to accelerated clearance of anti-TNF drugs [104, 105]. Disease severity is associated with higher levels of serum and mucosal TNF . If no clinical improvement occurs within 7–10 days or a clinical deterioration occurs, surgery must be considered . Current rates of colectomy in acute severe UC treated with infliximab are reported between 35 and 50% . In patients responding to infliximab, combination therapy with AZA can be considered for 6–12 months until durable remission is achieved. Both infliximab and AZA are suitable for maintenance therapy. Tacrolimus may be an alternative in specialized centers .
Both cyclosporin and infliximab can be used for rescue therapy. Current data suggest equal efficacy in preventing short-term and long-term colectomy, with similar failure rates . The value of sequential rescue therapy after nonresponse to either infliximab or ciclosporin remains controversial. Although clinical response with sequential therapy showed short-term response rates of 62% and colectomy rates of 28%, the intense immunosuppression leads to serious infections in 7% and the limited data so far do not allow clinical guidance . Sequential therapy should only be performed in specialized centers.
The care for patients suffering from moderate-to-severe UC remains challenging, despite decreasing morbidity and mortality over the past decades. 5-ASA, thiopurines, anti-TNF, and vedolizumab are well-established treatment regimens for induction and maintenance of remission in UC. 5-ASA is considered the standard in treating mild-to-moderate UC, eventually combined oral and rectal. Patients with inadequate response to optimized 5-ASA therapy will escalate to budesonide-MMX or oral prednisone. To induce remission in more severe disease, IV corticosteroids remain the first-line therapy, while cyclosporin and infliximab are the mainstay of rescue therapies in acute severe UC not responding to IV corticosteroids. The early detection of severe flares, more therapeutic options for medical therapy, better intensive care, and improved surgical techniques have led to a significant decrease in morbidity and mortality in acute severe UC. One of the most important factors remains the interdisciplinary teamwork between of gastroenterologist and experienced colorectal surgeons at any time during the clinical course in order to decide upon the best timing for colectomy, which is still required in 30% of the patients.
There are still many open questions and knowledge gaps to be clarified by future research. In the future, upcoming treatment options with different molecular pathways and different modes of actions will further increase the need for personalized medicine. A better understanding of pathophysiological processes, pharmacogenomics, and predictive markers for disease activity will help to identify subpopulations of UC patients who will benefit from tailored treatment regimens to individuals.
E.B. received consultant and/or speaker fees from Abbvie, Janssen, MSD, Norgine, Pfizer, Takeda. M.H.M. received consultant and/or speakers fees and/or grants from Abbvie, Janssen, Lilly, MSD, Pfizer, Takeda, UCB, Vifor. A.M.S. received consultant and/or speaker fees from the following companies: Abbvie, Janssen, MSD, Pfizer, Takeda, UCB, Vifor. F.S. received consultant and/or speaker fees from the following companies: MSD, Janssen, Pfizer, Takeda. G.V.A. received honoraria or grant support from Abbvie, Ferring, Janssen, Pfizer, Takeda. P.R. received consultant and/or speaker fees from Abbvie, and Amgen. D.L. received consultant and/or speaker fees from Abbvie, Biogaran, Biogen, Ferring, HAC-pharma, Janssen, MSD, Novartis, Pfizer, Prometheus, Roche, Takeda, Theradiag, Tillots. M.M. received consultant and/or speaker fees from Abbvie, Ewopharma AG, MSD, Takeda, and -Vifor.
There are no funding sources to declare.
E.B., M.H.M.: literature research and drafting the article. M.H.M., A.M.S., F.S., G.V.A., P.R., and D.L.: revising the article critically for important intellectual content.