The use of adalimumab in the management of refractory Crohn’s disease


Dr G.-T. Ho, Gastrointestinal Unit, Western General Hospital, Edinburgh EH4 2XU, UK.


Background Adalimumab is a humanized monoclonal antibody targeting tumour necrosis factor-α. Recent clinical trials have demonstrated its efficacy in Crohn’s disease; however, experience in clinical practice remains limited.

Aim To investigate the efficacy and safety of adalimumab in the clinical setting.

Methods The clinical outcomes of patients with medically refractory Crohn’s disease treated with adalimumab in the Western General Hospital Edinburgh, over a 3-year period (2003–2006), were studied.

Results Twenty-two (14 females; age at therapy: 32.6 years) patients were treated using an 80/40 mg induction regimen followed by fortnightly 40 mg treatment. All had proven refractory/intolerant to corticosteroids and immunosuppression. Twenty patients had had previous infliximab infusions – of these eight (36%), six (27%), three (14%) had previous infusion reactions, no response and lost response to infliximab, respectively. Over a period of 1.0 years (IQR: 0.62–2.5), Kaplan–Meier analyses showed that 68% (seven nonresponders) were in clinical remission and 67% (five surgery – discounting oral CD) avoided further surgery for active disease. 59% required dose escalation to 40 mg weekly (0.55 years; IQR: 0.22–1.4). Three (50%) primary nonresponders to infliximab achieved remission. Two patients developed serious infective complications and one patient developed lung cancer.

Conclusions Adalimumab is efficacious in refractory Crohn’s disease, with benefit observed in infliximab primary nonresponders. However, many patients require escalation of dosing regimen.


Monoclonal antibodies targeting antitumour necrosis factor (TNF) are highly effective therapies in Crohn’s disease (CD) and to a lesser extent in ulcerative colitis (UC).1–4 Use of infliximab, a chimaeric monoclonal anti-TNF antibody is established in the current management of inflammatory bowel disease (IBD). Recent controlled trials have demonstrated the efficacy and safety of adalimumab (Humira; Abbot Laboratories, Abbott Park, IL, USA), a fully humanized subcutaneous delivered monoclonal antibody (immunoglobulin G1) in CD. The CLASSIC-I (CLinical assessment of Adalimumab Safety and efficacy Studied as an Induction therapy in Crohn’s disease) showed significantly improved remission rates at week 4 using 160 mg/80 mg induction regimen (36% vs. 12%; P = 0.001) in moderate-to-severe CD patients naive to anti-TNF therapy.5 The CHARM (Crohn’s trial of the fully Humanised Antibody Adalimumab for Remission Maintenance) trial demonstrated superior remission rates in patients maintained on 40 mg weekly and fortnightly compared with placebo (remission rates of 36%, 41%, and 12%, respectively; P < 0.001 at week 56).6 This conclusion was also supported by the recently published CLASSIC-II study, with remission rates of 79% (40 mg fortnightly), 83% (40 mg weekly) compared to 44% in the placebo group (P < 0.05) at week 56.7 In the group of patients who had lost response or were intolerant to infliximab, the GAIN (Gauging Adalimumab efficacy in Infliximab Nonresponders) induction study showed that 21% of patients could achieve clinical remission at week 4 (by using the higher dosage induction regimen of 160/80 mg).8

The development of adalimumab and other alternative anti-TNF treatments, such as certolizumab pegol has undoubtedly increased the therapeutic armamentarium in IBD, in particular, in the difficult group of patients who have demonstrated loss of efficacy and intolerance in conjunction with infliximab use.9–11 Nevertheless, several issues including safety considerations have made the extrapolation of the recent data for the management of IBD in the direct clinical setting difficult.12 For example, the CLASSIC and CHARM studies included a wide spectrum of patients with CD, whereas at present, it remains likely that the anti-TNF therapies would be targeted and licensed in the use in patients with medically refractory/aggressive disease [refractory to corticosteroids, surgery and immunosuppression such as mercaptopurine (6-mercaptopurine)/azathioprine and methotrexate]. It remains pertinent that patients with stomas are also excluded from the clinical trials. In addition, studies on maintenance therapy data have been limited by the fact that no data beyond 12 months have been presented. Further problems to be taken into account in evaluation of the clinical trials include the definite short-term complications associated with biological therapy, and also the uncertainties of long-term complications. Recent open-labelled data on adalimumab have provided useful insights into the areas discussed.11, 13


We present the clinical data on the use of adalimumab with the primary emphasis on efficacy, safety, need for dose escalation, further surgery and cost, on the basis of experience at the Gastrointestinal Unit, Western General Hospital serving Lothian region (catchment areas 650 000 with tertiary referrals from across Scotland). Between September 2003 and August 2006, we treated 22 patients (14 females; median age 32.6 years; IQR: 20.9–41.7) with adalimumab using an 80/40 mg induction regimen followed by 40 mg fortnightly maintenance (Table 1). Nine patients were assessed from previous open-labelled adalimumab trial (Clinical study protocol M04-690/M06-829 Abbot – multicentre open-label study). The median follow-up period was 1.0 years (IQR: 0.62–2.5 years). Fourteen (63%), six (27%), one (5%) and one (5%) of patients had colonic, ileo-colonic, ileal and oral CD, respectively. Five (23%) patients also had concurrent perianal disease.

Table 1.   Demographics of adalimumab-treated cohort (n = 22)
Age at diagnosis (years)16.2 (IQR: 11.2–45.2)
Age at adalimumab therapy (years)32.6 (IQR: 20.9–41.7)
Median follow-up period (years)1.0 (IQR: 0.62–2.5)
Disease location
 Ileal 1 (5%)
 Colonic 14 (63%)
 Ileo-colonic6 (27%)
 Oral1 (5%)
 Perianal disease5 (23%)
Previous infliximab therapy20 (91%)
Infliximab reaction8 (36%)
No response6 (27%)
Loss of response3 (14%)
Long drug holiday/trial3 (14%)
Concomitant immunosuppression15 (68%)

We recorded the disease activity/extent, previous medical therapies, development of adverse events, need for further surgery (for active disease) and cost over the course of the follow-up. Clinical remission at time of latest follow-up was defined by the regression of symptoms, in concordance of both patient/doctor global assessments (and normalization of inflammatory markers). This was a retrospective case-note review study and Kaplan–Meier survival analyses were used to provide censored data for the outcome data above. All patients irrespective of previous operations, treatments and disease location were included in this study.


Patient selection/cohort

This current cohort comprised medically refractory patients (Table 2) where standard medical therapies involving infliximab, mercaptopurine/azathioprine, methotrexate and elemental dietary approaches had failed in controlling disease activity. 20 (91%) patients had been previously with infliximab. Of these, eight (36%), three (14%) and six (27%) patients were intolerant, lost response and failed to respond to infliximab, respectively. We define the loss of response to infliximab as initial response followed by subsequent reduction in the efficacy requiring a shortened period of re-treatment and eventual loss of response. All patients had more than five infliximab infusions but no patient underwent dose escalation to 10 mg/kg regimen. All three patients who had lost response were receiving infliximab infusions of <6 weeks in frequency with observed duration of effect between 2 and 4 weeks.

Table 2.   Details of current and previous medical therapies in adalimumab-treated cohort
NumberCaseCurrent therapiesPrevious failed/ intolerant therapies
  1. * Enrolled from open-labelled adalimumab trial.

  2. MP, mercaptopurine.

 1*62, F
Colonic CD
Oral methotrexate – ineffective as a single agent
Pentasa (mesalazine)
Allergic reaction to infliximab (2004)
Azathioprine – ineffective, discontinued
Corticosteroids – diabetes/hypertension
 2*39, F
Ileo-colonic CD
Azathioprine (2 mg/kg)Loss of response to infliximab (2003)
Multiple oral courses of corticosteroids
 3*21, F
Ileo-colonic CD
Pyoderma gangrenosum
SulphasalazineLack of response to episodic infliximab (2003)
Azathioprine – intolerant (2003)
Elemental diet (2003)
 447, F
Ileo-colonic CD
Ankylosing spondylitis
Oral methotrexateLoss of response to infliximab (2005)
Azathioprine/MP – significant neutropaenia
 541, M
Colonic CD
Episcleritis, arthropathy
MP (1 mg/kg)
Long-term oral prednisolone
Primary nonresponse to infliximab (2005)
Long-term oral prednisolone
Mesalazine (nausea)
Azathioprine (nausea)
 634, M
Ileal + Perianal CD
Previous ileal resection
Pentasa 2 g b.d.Reaction to infliximab (2005)
Azathioprine/MP (arthropathy)
Methotrexate (mucositis and liver dysfunction)
 758, M
Colonic CD
Azathioprine (2.5 mg/kg)
Oral prednisolone (>6 weeks)
Reaction to infliximab after drug holiday (2006)
 821, M
Ileo-colonic CD
Previous colectomy (ileostomy)
Azathioprine (2 mg/kg)Reaction to infliximab after drug holiday (2005)
Azathioprine (lack of efficacy)
Methotrexate (lack of efficacy)
 918, F
Colonic CD
Azathioprine (2 mg/kg)
Oral prednisolone
Primary nonresponse to infliximab (2005)
Multiple courses of oral steroids
Antibiotics – metronidazole
1020, F
Erythema nodosum
Azathioprine (1 mg/kg – neutropaenia)Reaction to infliximab (2005) following drug holiday
Azathioprine (lack of efficacy, neutropaenia)
Methotrexate (nausea)
Elemental diet (lack of efficacy)
1123, F
Ileo-colonic CD
Previous colectomy
Long-term low-dose oral prednisolone
Pentasa 1 g b.d.
Azathioprine/MP (lack of efficacy and hair loss)
Methotrexate (lack of efficacy and intolerance)
Elemental diet (lack of efficacy)
1235, F
Oral CD
 Infliximab – long drug holiday (2001)
Azathioprine (lack of efficacy)
Tacrolimus (lack of efficacy)
1343, F
Ileo-colonic CD
Perianal disease
Previous colectomy (ileostomy)
Oral methotrexateReaction to infliximab
Azathioprine (lack of efficacy)
1431, M
Colonic CD
Perianal disease
Azathioprine (2 mg/kg)
Low-dose long-term oral prednisolone
Reaction to infliximab
Azathioprine (lack of efficacy)
15*50, F
Colonic CD
 Reaction to infliximab
Azathioprine (lack of efficacy)
16*20, M
Colonic CD
Perianal disease
Oral prednisolonePrimary nonresponder to infliximab
Methotrexate (lack of efficacy)
Azathioprine (lack of efficacy)
17*28, F
Colonic CD
Perianal disease
Parenteral methotrexate (limited efficacy)Primary nonresponder to infliximab
Methotrexate (lack of efficacy)
Azathioprine (lack of efficacy)
Elemental diet (lack of efficacy)
18*35, M
Colonic CD
Parenteral methotrexate
Long-term oral prednisolone
Loss of response to infliximab (2004)
Azathioprine (lack of efficacy)
19*18, M
Colonic CD
Pentasa 1.5 g b.d.Loss of response to infliximab (2003–2004)
Parenteral methotrexate (lack of efficacy/ neutropaenia/nausea)
Azathioprine (lack of efficacy)
2024, F
Ileo-colonic CD
MP (1 mg/kg)Reaction to infliximab (2005)
Azathioprine (neutropaenia)
Elemental diet (lack of efficacy)
Oral methotrexate (lack of efficacy)
2119, F
Colonic CD
Oral methotrexate
Oral budesonide
Primary nonresponse to infliximab (2005)
Azathioprine (lack of efficacy)
22*45, F
Colonic CD
Erythema nodosum
Long-term oral corticosteroids
Azathioprine (2 mg/kg)
Pentasa 1 g b.d.
Azathioprine (lack of efficacy)

In addition to infliximab, five (30%) had been previously treated with azathioprine, mercaptopurine, methotrexate and elemental diet without success. Three patients had surgery (two colectomies, one small bowel resection) prior to initiation of adalimumab. Eight (36%) demonstrated corticosteroid dependency in the 6-month period prior to therapy.

Efficacy and need for surgery in nonresponders.  Using Kaplan–Meier analyses, 68% of patients were in clinical remission at 2 years censor (two patients were not included in further analysis: one patient withdrew immediately following severe injection pain and a further developed severe cellulitis – 7 and 33 days, respectively; Figure 1a). Sixty-seven per cent of patients avoided the need for further surgery (discounting the patient with oral CD) at 2 years censor (Figure 1b). Of those who have failed adalimumab, five (23%) patients underwent further surgery (median time after initiation of adalimumab – 0.63 years, IQR: 0.22–1.7; all with colonic CD requiring colonic resection). Further two patients were considered to have failed therapy (one considering surgery and a further, awaiting further trial of certolizumab/tacrolimus) at the time of latest follow-up. Two patients with ileostomy (previous panproctocolectomy) were maintained in clinical remission (range of follow-up: 0.40–0.62 years), but both required escalation to weekly adalimumab treatment. There were no significant differences in the inflammatory markers at the time of initiation and latest follow-up (CRP 15 vs. 8 mg/L, P = 0.6; albumin 37 vs. 39 g/L, P = 0.9 and white cell count 8.4 vs. 6.2 × 109, P = 0.6). Moreover, there were no differences in the efficacy or need for surgery in patients who were followed up in the trial setting, five (63%) of nine patients were in clinical remission at the time of late follow-up.

Figure 1.

 (a) Proportion of patients in clinical remission at time of latest follow-up. (b) Proportion of patients in clinical remission at time of latest follow-up. (c) Kaplan–Meier survival curve showing proportions of dose escalation from fortnightly 40 mg to weekly. The value presented in text, is the actual percentage of patient requiring dose escalation, i.e. 13 of 22 patients (59%) which mirrors the estimated censored data provided by  (c) at 6 months and 1 year.

Dose escalation to 40 mg weekly adalimumab.  It of interest to note that 59% of patients required dose escalation to 40 mg weekly treatment (median time of escalation – 0.55 years, IQR: 0.22–1.5; Figure 1c). At 6 months, censored data revealed that approximately 50% of patients on adalimumab monotherapy and on concomitant immunosuppressive therapy required dose escalation. Six (36%) of these patients had intervals of adalimumab-free period without loss of efficacy on recommencement (one patient with ileo-colonic CD were successfully re-treated following colectomy).

Primary nonresponders to infliximab.  The subgroup of primary nonresponders to infliximab was of particular interest. Three (50%) primary nonresponders to infliximab were successfully treated and maintained in clinical remission at latest follow-up (median 0.99 years, IQR: 0.77–1.5).

Patient A (19-year-old female with early-onset Crohn’s colitis and learning difficulties – treated with three infliximab infusions in May/July 2005 with no objective/subjective response), was given adalimumab in December 2006 with dose escalation to 40 mg weekly in March 2007 – concomitant oral methotrexate 20 mg and oral budesonide at the time of switch. Currently remains in clinical remission.

Patient B (31-year-old male with CD pancolitis, episcleritis and arthropathy) received two infusions of infliximab and maintained on long term of oral prednisolone in June 2005. Neither luminal nor extraintestinal manifestations improved and adalimumab was used in November 2006 with a good subsequent improvement in all symptoms and the withdrawal of long-term oral prednisolone.

Patient C (18 years old with early-onset colonic CD) received three infliximab infusions in July/August 2005 (also on concomitant azathioprine and long-term oral steroids) but failed to demonstrate improvement necessitating further high-dose oral prednisolone. Adalimumab was given in October 2005 with a good early response symptomatically mirrored also by a significant reduction in inflammatory markers, faecal calprotectin levels and withdrawal of oral corticosteroids. Adalimumab dosing regimen was escalated to 40 mg weekly by March 2006. Of the three patients who did not improve on adalimumab (from the primary nonresponders to infliximab) two patients required colectomy (2.5 and 0.6 years treatment, respectively).

Adverse events/safety

Overall, subcutaneous injections of adalimumab were well tolerated with no cases of anaphylaxis, acute or delayed hypersensitivity reactions, but there was one case of early withdrawal because of severe injection site pain.

There were three (14%) serious adverse events: two sepsis-related complications, colonic perforation leading to the development of psoas abscess and severe facial cellulitis. The case of colonic perforation developed in a 23-year-old male with CD colitis and perianal disease (previously treated with infliximab, azathioprine and methotrexate) treated with adalimumab 40 mg fortnightly (for 142 weeks/2.6 years). The development of the former complication was considered to be secondary to inadequate response to adalimumab in active CD. A colectomy performed in this patient confirmed active CD in the left colon. Severe facial cellulitis developed in a 35-year-old female with oral CD (previously treated with infliximab, azathioprine and tacrolimus) following 8 weeks of adalimumab therapy, requiring immediate discontinuation of treatment and in-patient management in the form of intravenous antibiotics.14

A case of locally advanced nonsmall cell lung cancer developed in a 70-year-old female (34 cigarette pack-years) with CD colitis. This patient had been treated concurrently with oral methotrexate (also previously treated with infliximab, parenteral methotrexate and azathioprine) and was maintained in clinical remission with 40 mg weekly adalimumab therapy (92 weeks/1.7 years). Therapy was withdrawn immediately in June 2006. The causal link with adalimumab or indeed other immunosuppressive agents used was considered to be uncertain, by respiratory doctors at this time, given the significant smoking history. Nevertheless, further respiratory/oncological review is in progress.

The current local costing for financial year 2005–2006, for adalimumab 40 mg fortnightly treatment was £10 773 (£21 546 for adalimumab 40 mg/weekly), compared to infliximab at 5 mg/kg dosing regimen 8-weekly maintenance therapy of £7112 (Source: Mr D. McCabe, Senior Pharmacist, Pharmacy Department, Western General Hospital).


We describe the first United Kingdom experience in the use of adalimumab in the real life clinical setting. Several findings from our data set extend the knowledge of assumed current practice beyond that provided by the available published clinical trials. First, we show that adalimumab can be efficacious in a patient cohort with medically refractory CD. A majority of the patients we described were previously treated with infliximab, or proven intolerant or resistant to most immunosuppressive therapies, and also with previous surgery – a perceptibly different spectrum of patients, but reflecting the reality of clinical setting. Notwithstanding this, 68% of these patients were maintained in clinical remission and 67% (discounting patient with oral CD) avoided the need for further surgery at the time of latest follow-up with a median follow-up period of 1.0 years. However, our follow-up period did not extend beyond that provided by the maintenance trials as highlighted earlier; but, this reflects the only recent introduction of adalimumab outwith the trial setting. This current data set, however, has now provided the platform to address the need for a larger register with longer follow-up data.

Secondly, we have also shown, for the first time to our knowledge, that the use of adalimumab can be of particular benefit in the primary nonresponders to infliximab. In this series, 50% of patients achieved and were maintained on clinical remission, having demonstrated no prior response to infliximab. The GAIN study and other published series have shown that alternate anti-TNF strategies (also certolizumab) can be useful in patients who have lost response or developed reaction to infliximab.8, 11 Our findings therefore advance this experience to the important subset of primary infliximab nonresponders, drawing similarity with the rheumatology experience, where alternative anti-TNF is commonly used interchangeably in the failure of a particular anti-TNF treatment. Further studies in primary nonresponse area are clearly relevant.

Our current practice of using 80/40 mg induction regimen followed by 40 mg fortnightly treatment was derived primarily from the rheumatology experience. Our data suggest that this protocol may require a re-evaluation in CD. Although both CLASSIC II and CHARM studies showed no significant difference between 40 mg weekly and fortnightly regimen, patient cohort/selection is clearly the critical determinant. As pointed out earlier, at present, the placing of the newer anti-TNF therapies is likely to be targeted at patients with severe medically refractory disease as described in this cohort. We show that nearly 60% of patients required dose escalation to 40 mg weekly within 6 months of therapy. Given that most dose escalation occurred early, the optimal dose induction regimen is also a subject of much debate whether 160/80 mg induction may be more appropriate. The efficacy of this higher induction regimen is currently under evaluation in the trial setting. Weekly adalimumab has substantial differential cost implications and this needs to be factored into the pharmaco-economic analyses for funding or regulatory bodies of healthcare provision, mostly relevantly in the United Kingdom. It is noteworthy to highlight also that Kaplan and colleagues recently showed that in patients who have lost response to infliximab, dose escalation will yield more quality-adjusted life-years compared to switching to adalimumab, however, at a considerable cost.15

The safety issues surrounding anti-TNF therapies (and other new biologics) are of great importance. The incidence of short-term serious adverse events seen in this small cohort is proportionate to that reported in clinical experience of infliximab in larger series and brings into sharp focus the paramount requirement for vigilance. Of the serious adverse events, only one case (severe facial cellulitis) was considered to be wholly attributable to adalimumab. It also remains pertinent, of course, that the longer term effects of adalimumab or indeed all the anti-TNF therapy are not fully known. The recent arguments for and against concomitant immunosuppression as a form of synergy with anti-TNF strategy with the aim to reduce the induction of anti-TNF antibodies are well summarized by a recent editorial by Hanauer et al.16 The above concept is again under re-evaluation because of the concerns of potential increase in life-threatening infections and malignancies, particularly in children.17 Of particular concern is the recent description of rare hepatosplenic T-cell lymphomas in CD treated with infliximab and concomitant azathioprine or mercaptopurine, with the immunosuppressive load the likely implicating factor.12 Accepting that the overall numbers are small, we have demonstrated no difference in the efficacy or need for escalation for adalimumab as monotherapy or with concomitant immunosuppression.

There are several limitations to this case-series, which merit further discussion. First, the data presented are based on a retrospective review and not based on a uniform fixed time-point of analysis (hence the use of Kaplan–Meier analysis). Secondly, the outcome data are not directly comparable to the clinical trials because of different measures for end-points (CDAI in the case of trials, and the exclusion of patients with stoma, recent intestinal resection for examples). Thirdly, the current data reflect our clinical practice, although a standardized approach is adopted and presented based on one-centre, with considerable experience in difficult IBD cases. It is of course noteworthy that it is our intention to present a data set, which most closely reflects the true clinical setting. We are in the process of establishing a national database of second-generation anti-TNF experience using the Scottish Society of Gastroenterology network. The potential importance of such register is highlighted by the success of the rheumatology practice in the form of the British Society for Rheumatology Biologics Register Control Centre Consortium, where continuous monitoring of new and potent therapies allows dynamic re-evaluation of clinical practice with safety being the critical issue.18 It is also pertinent that the current cohort was largely drawn up of patients with colonic involvement only. Whilst it is of interest to ascertain whether disease location may act as a determinant to therapeutic response, the overall size of this cohort is insufficient to permit valid sub-phenotypic analyses.

In conclusion, these data provide evidence that adalimumab is efficacious in many patients with otherwise medically resistant CD. Of particular promise is the novel observation that adalimumab appears to be efficacious in primary nonresponders to infliximab. However, a significant proportion of patients require adalimumab dose escalation. Serious short-term complications are to be expected in these patients after treatment, with long-term risk remaining unquantified, and the risk–benefit ratio that needs to be fully discussed with the patient.


We acknowledge Dr James Rose (Ayr Hospital), Dr Bill Ruddell (Falkirk Royal Infirmary) and Dr Peter Bramble (Stirling Royal Infirmary) as referring physicians to our cohort. Declaration of personal interests: G.-T. Ho, C. Lees and I. Arnott have received financial sponsorship (for educational and travel grants, and advisory fees) from Shire Pharmaceuticals, Schering Plough, Abbott Pharmaceuticals, Proctor and Gamble, UCB and Otsuka Pharmaceuticals. J. Satsangi has acted as a member of advisory boards and given lectures for Schering-Plough, Abbott, Ferring and UCB. Declaration of funding interests: None.