Use of infliximab in the prevention and delay of colectomy in severe steroid dependant and refractory ulcerative colitis

Abstract

AIM: To determine if infliximab can prevent or delay surgery in refractory ulcerative colitis (UC).

METHODS: UC patients who failed to have their disease controlled with conventional therapies and were to undergo colectomy if infliximab failed to induce a clinical improvement were reviewed. Patients were primarily treated with a single 5 mg/kg infliximab dose. The Colitis Activity Index (CAI) was used to determine response and remission. Data of 8 wk response and colectomy rates at 6 mo and 12 mo were collected.

RESULTS: Fifteen patients were included, 7 with UC unresponsive or intolerant to IV hydrocortisone, and 8 with active disease despite oral steroids (all but one with therapeutic dosage and duration of immunomodulation). All the IV hydrocortisone-resistant/intolerant patients had been on azathioprine/6-MP < 8 wk. At 8 wk, infliximab induced a response in 86.7% (13/15) with 40% in remission (6/15). Within 6 mo of treatment 26.7% (4/15) had undergone colectomy and surgery was avoided in 46.6% (7/15) at 12 mo. The colectomy rate at 12 mo in those on immunomodulatory therapy < 8 wk at time of infliximab was 12.5% (1/8) compared with 100% (7/7) in patients who were on long-term maintenance immunomodulators (P < 0.02).

CONCLUSION: Infliximab prevented colectomy due to active disease in immunomodulatory-naïve, refractory UC patients comparable to the use of Cyclosporine. In patients, however, on effective dosage and duration of immunomodulation at time of infliximab therapy colectomy was not avoided.

INTRODUCTION

Ulcerative colitis (UC) is characterized by a life-long chronic course with remissions and exacerbations, with 15% of patients having a severe attack requiring hospitalization at some time during their illness. While the management of a severe exacerbation has traditionally been dependent on intravenous steroids, up to 40% of patients fail to respond resulting in colectomy during that admission[1,2], this rate has remained unchanged over 30 years[3]. Even in those who respond, 25% become steroid dependent and an additional 20%-30% require colectomy at 12 mo[4]. Cyclosporine (CsA) has been shown to be an effective rescue therapy in acute UC in up to 80% of patients[5], but despite this and the use of azathioprine or 6-mercaptopurine (6-MP) as maintenance therapy, over 65% of patients have relapsed and 30% require colectomy within 12 mo[6–8]. Concerns over the safety profile of CsA, even with low dose CsA (2 mg/kg per day), has also resulted in reluctance for clinicians to use this due to its well documented side effects[9].

The early inflammatory bowel disease studies investigating infliximab, a chimeric monoclonal antibody to the proinflammatory cytokine tumour necrosis factor alpha (TNF-α), focused on Crohn’s disease where it was known that TNF-α was implicated. Initially the role of TNF-α was considered to be less predominant in UC although more recently TNF-α has been shown to play an important role in the inflammatory process in UC[10,11]. The initial uncontrolled published studies suggested efficacy in the use of infliximab for acute steroid refractory UC[12–14], although only two of three small randomised placebo controlled trials supported its role[15–17]. Two large randomised placebo controlled trials (ACT 1 and ACT 2) recently confirmed the efficacy of infliximab in moderate UC[18] and a recent meta analysis of randomised studies in UC demonstrated infliximab to be more effective than placebo, with a mean response rate of 68% and a 40% remission rate at 2 wk, giving the patient number needed to treat as 3 to 5[19]. In clinical practice, however, patients with acute severe UC requiring IV steroids and chronic steroid refractory UC are clearly distinguished but limited data exists on any differences in efficacy of infliximab in long term outcomes in either group[20,21]. Only the ACT 1 and ACT 2 studies provide data beyond 3 mo in a randomised controlled trial. These subjects achieved a 45% response and 35% remission at 54 wk (ACT 1), but patients received infliximab every 8 wk following a three dose induction at 0 wk, 2 wk and 6 wk and 78% remained on steroids at one year. These data do not reflect current practice in the management of most patients with UC in the majority of countries around the world and there is little data on long-term outcome following a single dose of infliximab in patients with UC not responsive to standard medical therapy.

In this paper we present our clinical experience of using infliximab in the context of patients with both acute severe UC and chronic refractory UC where surgery was the planned outcome if infliximab failed to induce a clinical improvement in either group.

MATERIALS AND METHODS

All subjects were patients at the Fremantle Hospital in the southern metropolitan region of Perth, Australia, which is a 450 bed tertiary institution and is a centre for the management of IBD. All patients receiving infliximab for IBD were considered for treatment only after failure of disease control with conventional therapy including 5-aminosalisylic acids, steroids and immunomodulation and were treated between August 2001 and December 2006. All patients were planned to go to surgery if treatment with infliximab failed. Patients received a single induction dose of 5 mg/kg in all but one case. Prior to the infliximab infusion patients received a 200 mg IV injection of hydrocortisone unless already on IV hydrocortisone. No prophylactic antibiotics were given. Tuberculosis was excluded in all patients both by a normal chest X-ray and a negative Mantoux test or QuantiFERON-TB Gold (QFT-G, Cellestis Limited, Carnegie, Victoria, Australia). Patients’ records were reviewed retrospectively and data was collected from the hospital patient IBD database which includes demographic details, disease duration and extent, endoscopy and medication history, dates, doses and indications for therapy, blood results and recorded adverse outcomes and surgery.

Patients were classified as having UC according to the “Montreal Classification” (a modification of the Vienna Classification)[22]. The diagnosis of IBD had to be definite, and was made in accordance with previously established international criteria[23] based upon clinical, endoscopic, histopathological, and radiological findings. The diagnosis of UC was exclusive of infective enterocolitis (excluded by stool microscopy and culture, bacterial and amoebic serology, acid-fast staining of biopsies and mycobacterial cultures), Behcet’s disease and microscopic colitis. Demography, disease status, infusion number, response and remission rates and adverse effects were recorded.

A clinical response was defined as a reduction of 4 or more points in the colitis activity index (CAI)[24] and remission was considered to be a sustained CAI of less than or equal to 4. This was determined at 2-4 wk and again at 8 wk after the initial infusion. For remission to be considered to have occurred, steroids needed to be successfully withdrawn within the 8 wk period following the infliximab infusion without recurrence of symptoms. Time till colectomy and current UC status was determined from review of the patients’ records and IBD database.

Adverse effects were analysed. Adverse effects identified as unlikely related, possibly related and probably related to infliximab use. Serious adverse effects are defined as any adverse drug experience occurring that results in death, life-threatening adverse event, persistent or significant disability/incapacity, required in-patient hospitalisation, or prolonged hospitalisation or congenital anomaly or birth defect.

RESULTS

Patient demographics

Fifteen patients (10 male, 5 female) aged 18 years to 59 years (mean age 39 years) received infliximab at 5 mg/kg. Seven patients had acute severe UC at the time of infliximab, six were not responsive to 5 d of IV hydrocortisone and one was steroid intolerant. All seven of these patients had been on concomitant azathioprine/6-MP for < 8 wk. Eight patients had active disease despite oral steroids and therapeutic dosage and duration of immunomodulatory therapy in all but one patient at the time of infliximab. Fourteen patients were receiving 5-ASAs, with one patient intolerant of this therapy (Table 1).

Table 1.

Medications taken by patients at time of infliximab treatment

Medication	UC patients (n = 15)
5-ASA
Current	93.3% (14/15)
Intolerant	6.7% (1/15)
Steroids
IV	40.0% (6/15)
PO	53.3% (8/15)
Intolerant	6.7% (1/15)
Immunomodulators
AZA/6MP	86.7% (13/15)
< 8 wk	53.3% (8/15)
≥ 8 wk	33.3% (5/15)
Intolerant	13.3% (2/15)
Tacrolimus	6.7% (1/15)
Methotrexate	6.7% (1/15)

86.7% (13/15) of patients had extensive UC, one patient left-sided disease and one proctitis as defined by the Montreal classification[22]. The average CAI was 13 (SEM ± 2.8, range 8-20; Table 2) and 84.6% of patients had a raised CRP (Median 42 mg/L, range 8-110 mg/L) pre infliximab.

Table 2.

Individual patient data of disease extent, CAI scores pre and post infliximab, steroid usage and therapeutic immunomodulation (therapy > 8 wk at ≥ 1.5 mg/kg of 6-MP) at time of infliximab and time to colectomy

	Sex	Age at diagnosis	Age at infliximab	Disease extent	CAI at infliximab	CAI at 8 wk	On steroids IV or oral	Therapeutic immuno-modulation	Response to infliximab	Time to surgery (mo)
1	F	37	59	Pan Colitis	14	4	Intolerant	No	Complete
2	M	25	26	Pan Colitis	14	0	IV	No	Complete
3	F	28	28	Pan Colitis	11	N/A	IV	No	Partial	1
4	F	45	49	Pan Colitis	14	4	IV	No	Complete
5	F	54	54	Pan Colitis	20	6	IV	No	Partial
6	M	24	28	Pan Colitis	13	4	IV	No	Complete
7	F	22	22	Pan Colitis	11	6	IV	No	Partial
8	M	17	18	Pan Colitis	13	0	Oral	No	Complete	26
9	M	56	58	Proctitis	12	11	Oral	Yes	Nil	8
10	M	54	55	Left sided	15	N/A	Oral	Yes	Nil	< 1
11	M	32	40	Pan Colitis	10	6	Oral	Yes	Partial	12
12	M	25	31	Pan Colitis	16	3 (on steroids)	Oral	Yes	Partial	5
13	M	27	28	Pan Colitis	14	9	Oral	Yes	Partial	9
14	M	32	38	Pan Colitis	8	3 (on steroids)	Oral	Yes	Partial	9
15	M	56	40	Pan Colitis	13	1	Oral	Yes	Complete	6

Efficacy at 8 wk

At 8 wk 86.7% (13/15) of patients had responded to infliximab with 40% (6/15) in remission (Tables 2 and 3) as determined by the CAI. All those patients considered to be in remission also had had their steroids withdrawn by the 8 wk assessment. Of the responders, all had extensive colitis, while the two non-responders either had proctitis or left-sided UC (Tables 2 and 3). The median CRP at 8 wk was 21.4 mg/L (range < 1-88 mg/L).

Table 3.

Remission and response rates at 8 wk to infliximab in the UC patients and colectomy rates at 6 mo and 12 mo

Disease type	Response	Remission	Colectomy	Colectomy
	wk 8	wk 8	≤ 6 mo	≤ 12 mo
E1 - proctitis	0%	0%	0%	100%
	(0/1)	(0/1)	(0/1)	(1/1)
E2 - left sided	0%	0%	100%	100%
	(0/1)	(0/1)	(1/1)	(1/1)
E3 - extensive	100%	46.20%	23.10%	46.10%
	(13/13)	(6/13)	(3/13)	(6/13)
Total	86.70%	40.00%	26.70%	53.30%
	(13/15)	(6/15)	(4/15)	(8/15)

Efficacy at 6 mo

At 6 mo 26.7% (4/15) of patients had undergone colectomy, 2 within the first mo of infliximab and 1 at 5 mo and 1 at 6 mo (Tables 3 and 4). Three of these patients were on therapeutic dosages and durations of immunomodulators at the time of the infliximab therapy (defined as > 8 wk duration of 6-Mercaptopurine at ≥ 1.5 mg/kg). One of the two patients who had a colectomy within 1 mo of infliximab was not on a therapeutic duration of immunomodulation, but had no active UC endoscopically post infliximab. She, however, suffered from recurrent massive acute colonic bleeding that required hospitalization on 2 separate occasions that included blood transfusions. This patient chose to undergo colectomy. Pathology of the surgical specimen identified mild active chronic colitis with no cause of the massive colonic bleeding identified. Despite this, surgery resolved the bleeding issue. The other patient had had no response to the infliximab.

Table 4.

Colectomy rates at 6 mo and 12 mo compared to clinical response to infliximab

Response	No surgery	Colectomy	Colectomy
	at 12 mo	≤ 6 mo	≤ 12 mo
No response	0%	50%	100%
	(0/3)	(1/2)	(2/2)
Partial response	28.60%	28.60%	71.40%
	(2/7)	(2/7)	(5/7)
Complete response	83.30%	16.70%	20%
	(5/6)	(1/6)	(1/5)
Total	46.70%	26.70%	57.10%
	(7/15)	(3/16)	(8/14)

The patient who underwent colectomy at 6 mo also had a complete response to infliximab therapy but was intolerant of azathioprine/6-MP, and had suffered from chronic ankle pain that commenced within minutes of the infliximab infusion and remained for the next 6 mo. The patient also developed pneumonia 2 wk after infliximab therapy and when he developed a significant flare of his UC, he decided to undergo surgery rather than further infliximab treatments.

Efficacy at 12 mo

By 12 mo a further 26.7% (4/15) of patients had undergone colectomy giving a 12 mo colectomy rate of 53.3% (8/15). All of these patients were on therapeutic dosages and durations of immunomodulators at the time of the infliximab therapy. Three were on azathioprine/6-MP and one was on oral tacrolimus due to intolerance to azathioprine/6-MP. Of the 4 patients who underwent colectomy within 6 mo to 12 mo 1 was for steroid dependent proctitis. Two patients had colectomy at nine mo and one at 12 mo for pan colitis. These 3 patients had an initial partial response to infliximab (Tables 3 and 4).

There was a significant difference between the 12 mo colectomy rate for those patients who were on therapeutic dosages and durations of immunomodulators compared with those patients on immunomodulatory therapy for < 8 wk at the time of infliximab therapy [100% (7/7) vs 12.5% (1/8), P < 0.02].

Long term follow up

Of the seven patients who avoided surgery at 12 mo one patient required colectomy at 26 mo for an acute flare but six remained off steroids and was maintained on azathioprine/6-MP with follow up averaging 25.8 mo (range 12-56 mo).

Side effects

One patient had an adverse side effect to infliximab with the development of arthralgia in his ankle within minutes of initiation of the infliximab infusion, which persisted for 6 mo. The same patient had a significant adverse event with the development of community acquired pneumonia 2 wk post infusion which was treated with standard antibiotic therapy with no long term sequelae. No post operative complications were recorded.

DISCUSSION

Infliximab has become an accepted treatment option following recent data on its efficacy in moderate to severe UC with a 35%-40% remission rate at 8 wk[19]. Five randomised double blind studies have compared infliximab with placebo for the treatment of UC[15–18] with two further randomised studies comparing infliximab with steroids[25,26]. Despite significant heterogeneity in the study populations and trial designs used, all but one study demonstrated a significant improvement in the disease treated with infliximab compared to placebo[16]. Apart from the ACT 1 and ACT 2 studies, however, all the studies have had limited follow-up of 3 mo or less and have not used maintenance infliximab therapy. The remission rates in ACT 1 using infliximab at 5 mg/kg were 38.8%, 33.9% and 34.7% at wk 8, 30 and 54 respectively; with 78% of patients still requiring steroids at 54 wk despite 8 wk infliximab. Given the financial implications of regular maintenance infliximab therapy and the lack of data comparing induction therapy alone with induction plus maintenance therapy in longer-term outcomes in UC our study retrospectively analysed the outcomes of all UC patients at our institution who were to undergo colectomy if infliximab therapy did not induce a clinical improvement. All patients who received infliximab had confirmed ongoing active disease despite steroids.

The 86% response rate and 40% remission rate at 8 wk in our patients is comparable with previous published studies. At 6 mo only 4 of 15 patients had required colectomy, 1 of whom was a complete responder at 8 wk, but who was intolerant of 6-MP and who declined further infliximab treatment following a flare due to its side effects. Another patient chose to undergo colectomy within 1 mo of infliximab for recurrent massive PR bleeding for which no cause could be identified but endoscopically her UC was in remission. By 12 mo 53% (8 of 15 patients) had undergone colectomy. Five of the seven patients who avoided surgery were in remission at 8 wk following a single infliximab infusion and the remaining 2 had an initial response but not remission. Six of the seven patients who avoided surgery had acute severe UC unresponsive to IV hydrocortisone and had been on 6-MP for less than 8 wk at the time of infliximab. The seventh patient was on high dose oral steroids and had also been on 6-MP for less than 8 wk. The efficacy of infliximab in these patients with more acute disease in our institution is in keeping with higher response rates to infliximab in studies of steroid refractory UC when compared with steroid dependent cases[27], although the data remains conflicting[20,28]. It would be anticipated, however, that a greater role of TNF-α in severe UC would result in a greater clinical efficacy of infliximab in this setting. The fact that colectomy was prevented at 12 mo in 7 of the 8 patients, (with one patient having surgery but without active UC) who were naive to immunomodulators (ie less than 8 wk of therapy), suggests that infliximab could be used as bridging therapy. This induces remission whilst giving more time for azathioprine/6-MP to reach therapeutic effect which can then be used to maintain this without the need for regular maintenance infliximab. A similar phenomenon has recently been reported with the use of CsA in acute severe colitis where the colectomy rate was significantly lower in patients who were commenced on azathioprine concurrently with CsA compared with those who were already on azathioprine[6–8]. In those patients where immunomodulatory therapy has failed, however, regular maintenance therapy with infliximab may be of greater benefit although further studies are required to assess the risks of regular biological therapy versus colectomy in this group of patients.

Our data suggest that at 12 mo the outcome following a single induction dose of infliximab is similar to that seen in the ACT studies, without the need for regular maintenance infusions, provided immunomodulatory therapy can be commenced and continued. This is supported by a recent paper from Oxford assessing their use of single dose infliximab in UC[20]. This has significant cost implications since the direct cost saving of single dose infliximab therapy with maintenance immunomodulators compared to maintenance infliximab therapy over one year would be €15 000 ($20 000 US) per patient. Whilst the limitations of this study, being a small, uncontrolled retrospective analysis, prevent direct comparisons being made, further controlled studies assessing the long-term outcomes of induction therapy without maintenance infliximab are warranted.

Infliximab was well tolerated in all but one patient who developed acute arthralgia, within minutes of the infusion, which lasted 6 mo. This patient also developed community-acquired pneumonia 2 wk after infliximab, which responded to standard antibiotic therapy. Despite complete remission at 8 wk he subsequently flared at 5 mo (having been intolerant of 6-MP) and went to colectomy after declining further infliximab therapy. The side effect profile of infliximab is highly relevant when considering rescue therapy for severe UC as both CsA and infliximab have similar response rates, although there is currently no randomised trial comparing their efficacy and safety in acute severe UC. Whilst the retrospective nature of our data analysis may not have detected mild adverse events associated with infliximab our data are consistent with published studies which demonstrate an adverse side effect profile in UC similar to that seen in Crohn’s disease[29]. This may favour the use of infliximab over cyclosporine as a rescue therapy, as there has been reluctance in some countries to use cyclosporine, even at 2 mg/kg, due to the risks of hypertension, seizures and nephrotoxicity[30].

In conclusion, infliximab induced a rapid response in over 80% of UC patients refractory to standard therapy with a 40% remission rate at 8 wk. There was avoidance of colectomy in 46% of all patients at 12 mo, without the need for regular infliximab maintenance therapy. There was, however, a significant difference in the 12 mo colectomy rate for those patients who were on therapeutic dosages and durations of immunomodulators compared with those patients on immunomodulatory therapy for < 8 wk at the time of infliximab therapy with 87.5% of these patients avoiding colectomy.

COMMENTS

Background

Corticosteroids have long been the predominant therapy for the induction of remission in severe ulcerative colitis (UC) with immunomodulators being efficacious in maintaining remission. However, over 30% fail to respond or loose response to therapy and require surgery. Cyclosporine has been used as a salvage therapy but 30% proceed to colectomy within 12 mo. The introduction of new biological agents such as infliximab have been shown to be efficacious in acute severe UC but data on long term follow up is limited.

Research frontiers

Infliximab has recently become an accepted treatment option in moderate to severe UC but the only prospective randomised long term data is in patients on maintenance infliximab for mild to moderate UC. This study was aimed at assessing the long term efficacy, safety profile and need for colectomy, following induction therapy with infliximab, without maintenance infliximab, in patients with UC who had failed standard medical therapy.

Innovations and breakthroughs

A single dose of infliximab had greater efficacy in patients with severe acute disease in our institution rather than in those with chronic UC who were flaring despite current treatment. This is in keeping with higher response rates to infliximab in studies of steroid refractory UC when compared with steroid dependent cases and may reflect a greater role of TNF-α in early disease.

Applications

Infliximab induction therapy is safe and appropriate in acute UC patients failing standard medical therapy but response rates appear poor in those with chronic refractory disease. A head to head randomised controlled trial of cyclosporine versus infliximab for the induction of remission in acute severe UC is urgently needed.

Terminology

Infliximab is a chimeric monoclonal IgG1 antibody directed against tumour necrosis factor-α (TNF-α) which is able to almost completely neutralize its biological activity.

Peer review

A single dose of infliximab had greater efficacy in patients with severe acute disease in this study rather than in those with chronic UC who were flaring despite current treatment. A small but useful retrospective study looking at the treatment outcomes following a single dose of infliximab as rescue therapy in refractory UC in a clinical practice setting.