Mentoring in IBD is an innovative and successful educational program for Canadian gastroenterologists that now includes an annual national meeting, regional satellites in both official languages, a website, an educational newsletter series, and regular electronic communications answering key clinical questions with new research. This issue is based on the presentation made by the contributing editors at the annual national meeting, Mentoring in IBD XVI: The Master Class, held November 6, 2015 in Toronto, Ontario.
Alessandro Armuzzi, MD PhD
The treat-to-target approach to managing patients with inflammatory bowel disease (IBD) involves risk-stratifying patients, treating high-risk patients with combined immunosuppression that incorporates biologic therapy, selecting a treatment target, implementing regular objective monitoring, and adjusting therapy to maintain the target (Figure 1).(1,2) Prospective studies are required to determine how this approach changes outcomes and patient quality of life, and whether treatment needs to be continued indefinitely in patients in prolonged deep remission. This important unanswered question has led clinicians, patients, and regulators to contemplate treatment discontinuation in this situation.
The major benefit of tumour necrosis factor (TNF)-α inhibitors is their effectiveness in moderate-to-severe and immunomodulator-refractory inflammatory bowel disease (IBD), with their potential for achieving mucosal healing and preventing disease progression and complications, including hospitalization and surgery. TNF-α inhibitors can significantly improve quality of life for patients in whom they are effective. Early combined immunosuppression also has the potential to alter the natural history of Crohn’s disease (CD).(3) Several real-world studies with long-term follow-up indicate that about two-thirds of patients treated with TNF-α inhibitors experience a sustained benefit from these therapies.(4–7)
Balanced against these benefits are the potential risks, including severe and opportunistic infections, malignancy, immune-mediated demyelinating neuropathies, cutaneous and allergic reactions, and infusion reactions. Concern is greatest for individuals at the extremes of age.
Data from the Therapy, Resource, Evaluation, and Assessment Tool (TREAT™) registry(8) and the Productivity, Safety and Efficacy: Long-Term Results in Adalimumab-Treated Patients with Crohn’s Disease (PYRAMID) registry(9) indicate an increased risk of serious and opportunistic infections with anti-TNF-α inhibitors in patients receiving triple immunosuppression that includes a corticosteroid, and with increasing age. A database study using the Food and Drug Administration Adverse Event Reporting System found an increased risk of lymphoma with TNF-α inhibitor use in combination with thiopurines.(10)
For many reasons—including inconvenience; dislike of taking drugs; fear of complications; and specific issues, such as pregnancy—patients may be reluctant to continue TNF-α inhibitor therapy indefinitely.
The Costs Of Inflammatory bowel disease in the Netherlands (COIN) study assessed the healthcare costs of IBD in a large cohort of patients (N=2252).(11) The study found that the overall cost of IBD is driven by medication costs, primarily by TNF-α inhibitor therapy, with hospitalization and surgery making up only a minority of healthcare costs. TNF-α inhibitors made up 64% of the cost of CD and 31% of the cost of UC. The cost of CD was almost three times higher than the cost of ulcerative colitis (UC).
The issues and unanswered questions described above have prompted an interest in the concept of de-escalating therapy or discontinuing TNF-α inhibitors in patients in remission, and several studies have addressed the impact of TNF-α inhibitor discontinuation on clinical status.
In the United Kingdom, regulators mandate reassessment of disease activity after 12 months of treatment with a TNF-α inhibitor and discontinuation in patients who have achieved clinical remission and mucosal healing. A retrospective audit of outcomes was performed in patients with IBD (N=160) in corticosteroid-free remission with a mean 25-month post-discontinuation follow-up.(12) Younger age at diagnosis (P=.003) and elevated white blood cell count (P=.044) predicted relapse, which was seen at 2 years in 56% of patients with CD and 50% of patients with UC or IBD unclassified. Reintroduction of TNF-α inhibitor therapy succeeded in recapturing response in 92% of patients.
A Spanish study evaluated outcomes in patients with IBD (N=1055) in clinical remission after discontinuing TNF-α inhibitor therapy.(13) Almost 50% of patients relapsed within 2 years. The risk of relapse was decreased in patients who continued an immunomodulator after relapse. The response rate to retreatment with the same agent was 88%.
The infliximab diSconTinuation in CrOhn’s disease patients in stable Remission on combined therapy with Immunosuppressors (STORI) trial evaluated outcomes and predictors of relapse after discontinuing infliximab in patients with CD (N=115) who continued immunomodulator therapy.(14) About 45% of patients relapsed after a median 28 months. Multivariate analysis found male sex, absence of surgical resection, white blood cell counts >6.0 x109/L, hemoglobin (Hb) ≤145 g/L, C-reactive protein (CRP) ≥5.0 mg/L, and fecal calprotectin ≥300 μg/g were predictors of relapse. Infliximab retreatment was effective in 88% of patients in the short term.
A systematic review of the impact of withdrawing biologic or immunomodulator therapy from patients with IBD in remission included 43 studies.(15) Analysis found that about 50% of patients who discontinued TNF-α inhibitor therapy after combination therapy relapsed after 2 years, with the proportion increasing over time.
A study investigating the effect of stopping TNF-α inhibitor therapy in patients (N=48) in long-term deep endoscopic remission included patients with detectable drug levels and those with undetectable drug levels.(16) During the first year after discontinuation, relapse occurred in 16 of 20 patients (80%) with detectable drug levels and in 9 of 28 patients (32%) with undetectable drug levels (P<.001). The finding of undetectable drug levels in patients with stable long-term remission may indicate that remission is no longer dependent on TNF-α inhibitor therapy.
A retrospective study evaluated outcomes in patients with CD (N=100) who discontinued infliximab after achieving clinical remission.(17) After a median 10-year follow-up, 52% of patients remained in sustained clinical remission. Most patients remained on immunomodulator therapy.
A retrospective analysis of treatment outcomes in patients with IBD (N=128) who restarted infliximab after a median 15-month discontinuation found that 70% maintained response at 1 year and 61% maintained response at >4 years.(18) Higher trough levels were associated with response and undetectable antidrug antibody (ADA) levels were associated with a reduced risk of infusion reactions.
Brian G. Feagan, MD FRCPC
Arguments can be made for discontinuing biologic therapy in patients responding to combination therapy, but well-designed withdrawal studies have not been conducted. Discontinuation may eradicate the treatment benefit in the very patients who generate the greatest morbidity and healthcare costs. Whereas reinitiation of treatment may recapture response in some patients, immunogenicity is an important concern, and high-quality data to address this issue do not exist. In addition, safety concerns with biologic therapy have been overstated.
Combination therapy with a TNF-α inhibitor and an immunomodulator is significantly more effective than immunomodulator therapy alone in patients with corticosteroid-dependent CD.(1) In addition, early use of combination therapy is significantly more effective than conventional therapy in inducing remission in patients with early CD.(2)
Data neither substantiate nor disprove the necessity of continued therapy for maintenance of remission, but several potential risks are associated with stopping therapy, including sensitization and the loss of an opportunity to change the natural history of the disease. Furthermore, stopping therapy may not necessarily be safer overall for the patient.
An azathioprine withdrawal study in patients with CD in clinical remission on long-term azathioprine (N=83) randomized patients to continued azathioprine or placebo for 18 months.(3) Relapse occurred in 3 patients in the azathioprine group (8% ± 4%) and 9 patients in the placebo group (21% ± 6%). Factors predicting relapse were baseline CRP >20 mg/L, time without corticosteroids <50 months, and Hb <120 g/L.
An azathioprine withdrawal trial in patients with CD treated with combination therapy (N=80) followed patients who continued or discontinued azathioprine for 2 years.(4) A similar proportion of patients in both groups required changes in the infliximab dose or discontinued infliximab during follow-up. Low patient numbers, however, prevent definitive conclusions from being drawn. Median CRP was significantly higher in the group that discontinued azathioprine (2.8 mg/L) than in the group that continued immunomodulator therapy (1.6 mg/L, P<.005). In addition, 5 to 15% of patients who discontinued azathioprine had undetectable infliximab trough levels after 1 year, compared with no patients in the continuation group.
A study of remission induction with infliximab plus methotrexate in 20 patients with early poor-prognosis rheumatoid arthritis found significantly reduced synovitis and erosions at 1 year, at which time infliximab was withdrawn.(5) At 2 years, treatment benefits were sustained in 70% of patients. However, one-half of patients required reintroduction of treatment, which was unsuccessful in 25%.
The principal investigator of the STORI trial, described in A European perspective (reference 14), commented, “Currently there is no good medical reason to stop IFX in patients in stable remission.” It is also important to realize that the predictors of relapse in this trial are associations, not causative factors. Logically, there is no good reason to suppose that IBD disappears with time in patients who have discontinued therapy as a result of being in remission. IBD develops as a result of the interplay of environmental and genetic factors that are not altered with short-term therapy, and long-lived pathogenic T lymphocyte clones are not ablated with conventional treatments. As a result, there is no theoretical basis for stopping biologic therapy.
Causality is difficult to establish when evaluating the incidence of serious infection in patients with IBD receiving multidrug therapy.(6) A Mayo Clinic case-control study of 100 consecutive patients with opportunistic infections and 200 matched controls assessed associations between drug therapy and infection. Multivariate analysis found that single-drug therapy was associated with an odds ratio (OR) for infection of 2.9 (95% confidence interval [CI]: 1.5–5.3), whereas use of 2 or 3 of drugs was associated with an OR of 14.5 (95% CI: 4.9–43). The relative risk of opportunistic infection was greatest in patients >50 years of age (OR: 3.0; 95% CI: 1.2–7.2).
Analysis of the adverse events in the Study of Biologic and Immunomodulator Naive Patients in Crohn’s Disease (SONIC) trial by drug therapy found a similar incidence of adverse events in patients in each therapy group: azathioprine, infliximab, and combination therapy.(7)
Analysis of safety data from the TREAT™ registry found a significantly increased risk of death with current use of corticosteroids (hazard ratio [HR]: 2.0, 95% CI: 1.3–3.0, P=.002) and narcotics (HR: 2.1, 95% CI: 1.3–3.2, P<.0001), and a significantly increased risk of serious infection with current use of corticosteroids (HR: 2.0, 95% CI: 1.4–2.9, P<.001) and narcotics (HR: 2.2, 95% CI: 1.5–3.1, P<.0001).(8,9) Risks of these two outcomes were not increased with infliximab or immunomodulators.
A pooled analysis of adverse events in randomized controlled trials of infliximab in IBD found no significantly increased risks of serious infection in CD, UC, or IBD as a whole.(10)
A prospective observational cohort study of the lymphoma risk associated with thiopurine therapy from the CESAME group included 19,486 patients with IBD.(11) On multivariate analysis, the HR for lymphoproliferative disorder for patients receiving thiopurines and those who had never received them was 5.28 (95% CI: 2.01–13.9, P=.0007) (Figure 2).
Discontinuation of biologic therapy, the most effective therapy for CD, carries a cost of relapse in a proportion of patients. No high-quality randomized controlled trials have examined the question of discontinuation in CD, and the therapeutic index of stopping therapy is unknown. No good reason for discontinuing therapy exists, and it is appropriate to continue therapy until good evidence for stopping exists.
A 32-year-old female with proximal jejunal Crohn’s disease measuring over 30 cm and perianal fistulous disease has been receiving therapy with a tumour necrosis factor (TNF)-α inhibitor plus methotrexate for 6 years. Clinically and biochemically her disease is in remission. The most recent computed tomography enterography, performed 4 months ago, shows no active disease in either the jejunum or the perianal region. The patient has had 4 miscarriages in the past and wants to get pregnant. She tells you that she is going to stop both the TNF-α inhibitor and methotrexate 6 months before attempting conception and throughout the pregnancy. This decision, made by her and her partner, is firm. She is, however, concerned about the risk of antidrug antibodies (ADAs) developing and wonders whether she should take azathioprine or mesalamine during her pregnancy. She also wants to know whether she would be able to restart therapy after the pregnancy.
As she has high-risk disease, stopping therapy could threaten her clinical status. It is appropriate to stop methotrexate prior to conception, but there is no reason to stop the TNF-α inhibitor. This patient may have made her decision to discontinue all treatment without having all the relevant information. It is therefore critical that she understand the data and the potential consequences of stopping therapy: the data on TNF-α inhibitor efficacy, her individual risk of relapse (40–50% during the first year), TNF-α inhibitor safety in pregnancy, and the risk to the fetus of flare and how a flare during pregnancy could be treated. Mesalamine would be unlikely to have any clinical benefit in this patient.
If she continues to insist she wants to stop her therapy, it is important to ensure that she is in endoscopic remission. Close monitoring of her clinical status would be indicated, including laboratory investigations and ultrasound. However, it is only worth monitoring her status if she is willing to accept treatment if signs of disease are found. In that situation, it would be important to restart therapy before full-blown relapse develops. Low infliximab levels on therapeutic drug monitoring (TDM) would indicate a greater likelihood of maintaining remission than high drug levels. Similarly, low ADA levels would indicate a greater chance of successful reinitiation of therapy after the pregnancy. It would be important to restart infliximab after delivery, as it might not be possible to recapture response if she relapsed. Azathioprine does not appear to increase the risk of congenital malformation, but taking it during the pregnancy may reduce the risk of development of ADAs and infusion reactions when she restarts infliximab.
A 52-year-old male with ulcerative colitis (UC) has been in remission on combination therapy with a TNF-α inhibitor and azathioprine for 6 years. He wants to stop the TNF-α inhibitor and is wondering about stopping all therapy. He asks you how long he is likely to stay in remission, if he would be able to restart the same TNF-α inhibitor if he had a flare, and if he needs any tests before stopping the TNF-α inhibitor.
Data are available to indicate relapse rates of approximately 18% per patient-year. After 1 year, 35 to 40% of patients relapse, and after 3 years, more than 50% relapse. No studies address relapse rates over longer time periods. No evidence addresses the risk of relapse for a patient with UC in long-term remission who stops all therapy. The severity of UC does, however decrease over time in some patients. Close monitoring would be appropriate, so that the TNF-α inhibitor could be restarted before a severe relapse developed.
Before stopping therapy, it would be appropriate to perform an endoscopy and biopsy to help gauge his risk of relapse. TDM would also be useful, as low drug levels could indicate a greater chance of maintaining remission. 6-thioguanine levels would also be useful. If he stops the TNF-α inhibitor, it would be wise to continue azathioprine, as this should help reduce ADA formation and the risk of infusion reactions when he needs to restart therapy. The best choice in that case would be the same TNF-α inhibitor he was originally taking.
John K. Marshall, MD MSc FRCPC AGAF, Chief of Gastroenterology Clinical Service, Hamilton Health Sciences; Professor of Medicine, Division of Gastroenterology, McMaster University , Hamilton, ON
Richard N. Fedorak, MD FRCPC FRCP (London) FRCS, Dean, Faculty of Medicine & Dentistry; Professor of Medicine, Division of Gastroenterology, University of Alberta, Edmonton, AB
Khursheed N. Jeejeebhoy, MD FRCPC PhD, Professor Emeritus, Department of Medicine, Department of Nutritional Sciences, University of Toronto, St. Michaels Hospital, Toronto, On
Alain Bitton, MD FRCPC, McGill University, Montreal, QC
Brian Bressler, MD MS FRCPC, University of British Columbia, Vancouver, BC
Anne M. Griffiths, MC FRCPC, University of Toronto, Toronto, ON
Steven E. Gruchy, MD MSc FRCPC, Dalhousie University, Halifax, ON
Remo Panaccione, MD FRCPC, University of Calgary, Calgary, AB
Craig Render, MD FRPCP, University of British Columbia, University of Alberta, Kelowna, BC
Hillary Steinhart, MD MSc FRCPC, University of Toronto, Toronto, ON
Jennifer Stretton, ACNP MN BScN, St. Joseph’s Healthcare, Hamilton, ON
IBD Dialogue 2016·Volume 12 is made possible by unrestricted educational grants from…
1B-391 Berkeley Street, Toronto, ON M5A 2X8
© Catrile & Associates Ltd., 2016. All rights reserved. None of the contents may be reproduced in any form without prior written permission from the publisher. The opinions expressed in this paper are those of the authors and do not necessarily reflect the opinions or recommendations of the sponsors, the grantor, or the publisher.