Background

Treatment options for steroid refractory acute severe ulcerative colitis (ASUC) are limited to infliximab (IFX) or cyclosporine and are further limited in patients with previous failure or adverse reactions to anti-tumor necrosis factor (anti-TNF) agents [1]. The colectomy rate in ASUC is 30% at 3-months and 40% at one-year post-discharge [2, 3], which underscores the refractory nature of this condition and highlights the need for research into alternate treatment options.

In the post hoc analysis of OCTAVE, clinical symptoms were reduced by day three of tofacitinib in patients with moderate-severe ulcerative colitis; however, hospitalized patients with ASUC were excluded [4]. Tofacitinib was an effective rescue therapy for ASUC in five small case-series limited by a short follow-up period [5,6,7,8,9]. Given the paucity of data, we aimed to evaluate the effectiveness of tofacitinib in hospitalized patients with ASUC in a real-world setting.

Methods

We retrospectively examined the charts of adult (> 17 years old) patients hospitalized with acute severe ulcerative colitis (ASUC) who were treated with tofacitinib between January 1, 2019 and September 1, 2020 at five centers across Canada. The centers included were McGill University Health Center (Montreal, Canada), Pierre Boucher Hospital (Longueuil, Canada), The Ottawa Hospital (Ottawa, Canada), Hamilton Health Sciences (Hamilton, Canada), and St. Paul’s Hospital (Vancouver, Canada). Patients were identified by physician recall. Patients were eligible if they: (1) had an established diagnosis of ulcerative colitis (UC) by standard clinical, endoscopic, and histological criteria; (2) were hospitalized with ASUC, defined as UC flare requiring hospitalization and IV corticosteroids as determined by physician judgment; and (3) received and were refractory to standard of care therapy as per the Canadian guidelines for ASUC [10]. The current guidelines recommend either infliximab or cyclosporine in individuals in whom parenteral steroids have failed after 72 h. Patients who were refractory to standard of care therapy were offered colectomy or tofacitinib in a shared decision-making process [10]. Induction doses (10 mg twice daily [BID] or 10 mg three times daily [TID] followed by 10 mg BID) were at the discretion of the treating physician. Patients with active C. difficile, enteric infection or cytomegalovirus infection were excluded.

The following data were collected: age, gender, disease duration, disease extent (Montreal classification), previous medical therapy, clinical status on arrival and during follow-up using the partial and full Mayo score, C-reactive protein (CRP), fecal calprotectin (FCP), hemoglobin, leukocytes, albumin level, and endoscopic assessment based on the Mayo endoscopic score.

Outcome Measures

The primary outcome was defined as the colectomy-free rate at 6 months. Secondary outcomes included clinical response rate (assessed at 7 days, 30 days, 90 days, and 6 months), hospital readmission rate, steroid use, and drug adverse event rate (thrombotic events, infection, and major adverse cardiovascular events (MACE)). A positive clinical response was defined by a decrease in the partial Mayo score ≥ 3. Clinical remission was defined as a partial Mayo score of ≤ 1. Normal biomarkers were defined as either normal CRP (≤ 5 mg/L) or normal FCP (≤ 250ug/g). Endoscopic assessment was collected if available. Endoscopic remission was defined as a Mayo endoscopic subscore 0 or 1.

Results

Eight patients with ASUC were included. All patients failed or were intolerant to anti-TNF therapy prior to initiating tofacitinib: three patients initiated anti-TNF therapy during current admission and five patients had anti-TNF exposure prior to hospitalization. All patients received corticosteroids prior to initiating tofacitinib. All three patients who initiated anti-TNF therapy during current admission received IFX at an induction dose of 10 mg/kg. All patients had severe symptoms, with a median partial Mayo score of 7.5 (IQR 6.25–9), as well as persistently elevated CRP that were refractory to standard rescue therapy. All patients received VTE prophylaxis during their hospitalization. No patients received prophylaxis against Pneumocystis jiroveci Pneumonia infections.

As induction, five patients received tofacitinib 10 mg BID while three patients received 10 mg TID for 3 days followed by 10 mg BID. Patients’ baseline characteristics are shown in Table 1. The median time between admission and tofacitinib initiation for patients that received sequential corticosteroids and IFX was 17 days (IQR 16–20.5); the median time for patients who received corticosteroids only prior to tofacitinib was 3 days (IQR 3–13).

Table 1 Patients baseline characteristics
Full size table

Patient outcomes are shown in Table 2. In 5/8 (62.5%) patients, clinical response was observed by day 7 after tofacitinib initiation. Among responders, the median time from tofacitinib initiation to discharge was 5 days (IQR: 5–6) and the median decrease in CRP was 64.2 mg/L (IQR:42.8–143.0) (See Supplementary, Appendix Fig. 1). At 30 days, all five responders were in clinical remission with a median prednisone dose of 25 mg per day (IQR: 21.3–25). At 90 days, all five responders were in steroid-free clinical remission with normal biomarkers. At 6 months, 3/5 responders remained in steroid-free clinical remission while 2/5 responders were readmitted for parenteral corticosteroids but did not require colectomy.

Table 2 Clinical response and Outcome in Patients with ASUC Treated with Tofacitinib
Full size table

At the last follow-up to-date, at 6 months following tofacitinib initiation, 5/8 patients were colectomy-free; 4/8 patients remained on tofacitinib; one patient switched to Ustekinumab. Three patients were refractory to tofacitinib and underwent colectomy; surgery was performed during the index hospitalization in two patients and at 30-days for the third patient. No postoperative complications were observed. During the follow-up period, 4/5 responders underwent repeat endoscopy: two patients had complete endoscopic healing (1 at 90 days; 1 at 6 months); two patients had endoscopic Mayo 3 colitis in the context of relapse.

Clinical response was observed in all three anti-TNF naïve patients who received tofacitinib after sequential corticosteroids and IFX. In contrast, clinical response was observed in only 2/5 patients with pre-admission anti-TNF exposure. Clinical response was observed in 3/5 patients who received 10 mg BID and in 2/3 patients who received 10 mg TID. No adverse were reported, including venous thromboembolism, infections requiring hospitalization, zoster reactivation, or major adverse cardiac events.

Discussion

In this multicenter case-series, tofacitinib administered to hospitalized patients with ASUC refractory to steroids and IFX was safe and led to a rapid clinical response in 5/8 cases, supporting the findings of five other retrospective case-series [5,6,7,8,9]. Additionally, the 6-month colectomy-free rate with tofacitinib (5/8, 62.5%) is comparable to the long-term outcome of patients treated with IFX and cyclosporine in the CONSTRUCT trial [2]. However, a subset of patients may experience a loss of efficacy over time, as 2/5 of our patients required parenteral corticosteroids at 6-month and 2/7 patients in Honap and colleagues required colectomy by 6 months [6]. We observed a lower effectiveness with tofacitinib among patients with pre-admission anti-TNF exposure, suggesting there may be a different efficacy based on the disease phenotype. In addition, some patients received an induction dose of 10 mg TID, an off-label dose. The effectiveness is similar between different induction doses within our cohort. However, both observations must be cautiously interpreted due to the small case numbers. Lastly, the patients who receive sequential IFX and tofacitinib had significant of immunosuppression and larger studies are needed to elucidate the safety of such sequential immunosuppression prior to the widespread adaptation of this practice. The necessity of prophylaxis against opportunistic infections, such as Pneumocystis jiroveci Pneumonia, should also be further studied.

A post-marketing study in patients with rheumatoid arthritis and cardiovascular risk factors reported higher rates of venous thromboembolism, cancer, and major adverse cardiovascular events with tofacitinib 10 mg BID compared to anti-TNF therapy. Although no significant adverse events occurred in our case-series, long-term safety remains to be seen given the small sample size in our study.

Tofacitinib may be an effective alternative for patients with refractory ASUC who are resistant to colectomy and provide time for decision-making and surgical planning. However, the evidence for its administration remains largely observational and limited by small sample size. Randomized controlled trials are needed to fully establish the efficacy and safety of tofacitinib in ASUC.