Objective Inflammatory bowel disease (IBD) is increasingly managed with the use of biologic therapies. National guidelines (National Institute for Health and Care Excellence (NICE)) suggest considering cessation after 1 year of therapy but lack detailed criteria for this. We aimed to describe clinical outcomes from the introduction of a biologic review panel (BRP) to implement modified criteria for cessation of antitumour necrosis factor (anti-TNF) therapy and step down to single-agent immunomodulator.
Design Retrospective review of patient outcomes following BRP implementation.
Patients All patients on biologic therapy discussed in the BRP within a 5-year period.
Setting Single IBD network covering three hospital sites.
Interventions Modified criteria for biologic cessation were based on published evidence; they excluded individuals with no suitable maintenance immunomodulator, previous surgery or evidence of active disease, additional indications for anti-TNF therapy and previous relapse on biologic cessation. All patients with IBD on a biologic were discussed at the BRP.
Main outcome measures Relapse following IBD cessation and relative cost of BRP.
Results 136 patients with IBD were reviewed, with 45 patients meeting the NICE guideline criteria for cessation. The BRP and modified criteria affected decision to withdraw therapy in 38% of these. Therapy was withdrawn in 27 patients, with a 20% 24-month relapse rate. Younger age at cessation was significantly associated with relapse (p=0.01).
Conclusion The BRP approach has proved a safe and effective means of decision making in stopping biologic therapy. Future work to inform exclusion criteria is required.
- biologic therapy
- inflammatory bowel disease
- health economics
Statistics from Altmetric.com
Significance of this study
What is already known on this topic
Antitumour necrosis factor agents can safely be withdrawn in some individuals with inflammatory bowel disease (IBD), but high relapse rates of up to 50% have been reported.
What this study adds
Detailed review of cases, ensuring deep remission and suitable immunomodulator therapy result in an acceptably low relapse rate in our experience.
How might it impact on clinical practice in the foreseeable future
More detailed criteria for cessation of biologic therapy in IBD may enable greater number of patients to safely stop therapy once in deep remission.
Inflammatory bowel disease (IBD) is often managed with immunosuppressant therapies. While thiopurines have traditionally been the maintenance agent of choice for Crohn’s disease (CD) and ulcerative colitis (UC) not controlled with 5-ASA compounds, antitumour necrosis factor (anti-TNF) agents and other biologics are increasingly being used1 for induction therapy and maintenance therapy. Although these agents are effective, they are costly and have potentially serious side effects.2–4 In addition, antibodies to the agents may develop which reduce their efficacy.5 Therefore, in individuals in steroid-free remission, current UK National Institute for Health and Care Excellence (NICE) guidelines, published in 2010, suggest planned cessation of therapy after 1 year unless there is clear evidence of ongoing disease activity. The guidelines state ‘Infliximab should be given as a planned course of treatment until treatment failure (including the need for surgery) or until 12 months after the start of treatment, whichever is shorter. People should then have their disease reassessed to determine whether ongoing treatment is still clinically appropriate’ (NICE technology appraisal TA187).6 No specific definition of remission is stipulated; a Crohn's Disease Activity Index (CDAI) less than 150 is given as an example of clinical remission. However, previous follow-up studies of such individuals have shown a relapse rate of approximately 40%–50% at 24 months.7–9 Other analyses have highlighted factors which predict risk of relapse post anti-TNF cessation, including male sex, age at diagnosis, smoking status, the absence of surgical resection, leucocyte count >6.0×109/L, haemoglobin ≤145 g/L, C reactive protein ≥5.0 mg/L and elevated faecal calprotectin. These are highlighted in the recent European Crohn's and Collitis Organisation (ECCO) guidelines.7 10–12 Additionally, specific features of Crohn’s disease such as colonic involvement, stricturing disease13 and perianal fistulation11 predict a lower likelihood of maintained remission. While the majority of those who relapse will respond to a further course of treatment with the same agent, a recent multicentre study reported that 12% lose their response on reintroduction of the same agent.13
We have previously described how the use of panel discussion in the form of an IBD multidisciplinary team meeting (MDT) can enhance biologic therapy decision making.14 Reviews of the use of MDTs in chronic disease management have suggested that further evidence of effectiveness of this approach is required.15 Here we detail the impact of an IBD biologic review panel (BRP) and evidence-based exclusion criteria for anti-TNF cessation on clinical outcomes for those stopping anti-TNF therapy.
Materials and methods
Since 2012, demographics and outcomes from all patients attending our IBD service treated with biologic therapy have been recorded in an IBD database to standardise clinical data collection and follow-up. The IBD BRP was set up to review the cases of all individuals on biologic therapy on an annual basis. For this study, all individuals remaining on anti-TNF therapy for over 6 months were evaluated in the BRP regarding suitability for de-escalation of therapy. For these patients, disease activity was assessed using clinical information (resolution of diarrhoea, abdominal pain and fatigue), radiological (where appropriate), biochemical (C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), albumin, calprotectin) and endoscopic tests. All those thought to be in remission were included for consideration of cessation of biologic therapy.
The NICE guidelines do not give specific criteria for remission but suggest this consists of ‘clinical symptoms suggesting absence of active disease and biological markers suggestive of remission with endoscopy if necessary’, suggesting a CDAI less than 150 for an example of clinical remission. It should be noted that the NICE guidance does not specify thresholds for biological markers of inflammation and does not mandate confirmation of endoscopic remission in the absence of other markers of activity. As recent evidence has suggested that those with histological remission have lower relapse rates, the BRP defined remission as the following:
No active clinical symptoms (ie, stool frequency <4 per day, no blood per rectum, no abdominal pain, no other symptoms likely to reflect active disease, eg, perianal symptoms).
Quiescent disease macroscopically at endoscopy (ie, no evidence of inflammation, eg, Mayo score 0 for UC or Simple Endoscopic Score for Crohn's Disease (SES-CD) <3 for CD) with quiescent or mildly active histological disease.
Radiologically no evidence of active inflammation for patients with Crohn’s with small bowel disease.
Biochemical evidence of resolution included a normal ESR and/or CRP, and where available a faecal calprotectin <250 μg/g. In some individuals where other causes of raised inflammatory markers were present, for example, active arthropathy, these were not used in determining whether the patient was suitable for drug withdrawal.
Use of the database and review panel helped identify any individuals who had not undergone complete evaluation at this time point, and further investigations were ordered as appropriate prior to reviewing the case.
All of those judged to be in deep remission according to the criteria above were discussed in the review panel to determine their suitability for anti-TNF cessation. The additional exclusion criteria for stopping anti-TNF agents applied are as follows:
No suitable immunomodulator: all patients were planned to step down to single-agent immunomodulator; therefore, those without an option for maintenance immunosuppression were excluded.
Postsurgical recurrence of IBD despite immunomodulator.
Difficult to control the disease, for example, requiring switch of anti-TNF therapy or frequent courses of steroids on single-agent immunomodulator.
Penetrating disease or perianal disease requiring anti-TNF therapy to achieve resolution.
Additional indication for anti-TNF.
Previous relapse on anti-TNF cessation.
The BRP consisted of a minimum of one gastroenterology consultant and two IBD nurse specialists, with input from a consultant radiologist, pathologist and surgeon available if required through a separate IBD MDT. In addition to the exclusion criteria listed, the MDT took a holistic approach to the patient, the course of their disease and other comorbidities which may have implications for therapy cessation, for example, rheumatological disorders.
Following cessation of therapy, patients were followed up for a minimum of 12 months. During this period, individuals had regular clinical assessments, as well as access to an IBD nurse telephone and email helpline. Relapse was defined as an increase in symptoms with appearances of active disease on biochemical, endoscopic and/or radiological testing requiring steroid therapy, surgery or recommencement of a biological agent. Those with clinical and biochemical/endoscopic confirmation of relapse were offered further treatment with the same anti-TNF agent in the first instance unless clinically contraindicated.
Between January 2012 and 2017, a total of 136 individuals underwent BRP review. Eighty-eight individuals received treatment with adalimumab and 51 with infliximab, with 17 receiving both therapies during this time period (see table 1 for demographics). Of these 45 (33%) met the NICE criteria for deep clinical remission. From these individuals, six (4%) were excluded as further investigations requested by the BRP or review of imaging revealed evidence of active disease. One additional patient moved out of the region during the BRP process. A further 11 (8%) did not meet the extended criteria for anti-TNF withdrawal. The number of patients excluded for each criterion is given in table 2. During this time period, 21 additional patients had biologic therapy withdrawn due to side effects, non-compliance or surgical resection of active disease.
Following the BRP process, 27 individuals (20%) were approved for planned cessation of anti-TNF therapy. This decision was communicated to the patient’s usual gastroenterologist, who then discussed drug cessation with the patient. No patients objected to drug withdrawal. A flow chart detailing the stages of consideration for cessation of anti-TNF therapy and exclusions is shown in figure 1. Following biologic cessation, these individuals were followed up for a median of 29 months (range 12–60 months). During this period a total of six individuals experienced relapse (22%). Demographics and clinical features comparing those relapsing with those who did not relapse are shown in table 3. The median time to relapse was 15 months (4–36 months; see figure 2), with a 1-year relapse rate of 11% and a 2-year relapse rate of 20% (4 of 20 patients followed up for at least 24 months). No clinical tests were significantly associated with relapse. Fewer than half of individuals provided samples for faecal calprotectin prior to cessation. In those where precessation calprotectin was available, there was no association with relapse; however, the lack of data makes it difficult to draw any firm conclusions from this. Relapse rates were equal in those taking infliximab at the time of anti-TNF cessation (4/18, 22%) and adalimumab (2/9, 22%). Younger age at cessation was significantly associated with a higher risk of relapse (p=0.01).
All relapsing patients suffered from Crohn’s disease, with none of the seven patients with UC experiencing relapse. This gives an unprecedented relapse rate of 0% in UC sufferers and a relapse rate of 30% in those with CD.
All six relapsing individuals were given anti-TNF agent as soon possible after symptoms were reported. The median time from diagnosis of relapse to first dose infusion was 105 days (range 28–300, the highest value being due to delay related to pregnancy). One relapsing individual required hospital admission, with the remainder managed solely in the outpatient setting. Four individuals were re-treated with the same agent. One patient switched agent as drug levels had been undetectable in the presence of antibodies prior to cessation, and the second switched agent due to preferred mode of administration. All individuals achieved remission, with only one requiring a course of steroids in the follow-up period (due to postpartum flare) and none required a further change of biologic agent.
Our data demonstrate the importance of careful patient selection when considering cessation of anti-TNF therapy. The 22% relapse rate seen in this series is one of the lowest reported. Use of the modified criteria and BRP discussion affected management in 17 of 45 (38%) of those meeting the NICE criteria for biologic withdrawal in this cohort. Where the decision was taken not to de-escalate patient therapy, this was most often because they were not in deep remission or because they had no suitable immunomodulator for maintenance therapy. It is possible that some individuals excluded would have been able to stop therapy without experiencing a disease flare. Further observational studies are required to determine the optimum criteria for drug cessation which allow the maximum possible number of individuals to stop treatment without relapse.
No statistical differences were observed in baseline laboratory parameters of those who relapsed compared with those who did not. There was also no association between the type of anti-TNF agent (prior to cessation) and relapse. While this may reflect the small numbers in this group, it suggests better biomarkers are required to predict the risk of recurrence of active disease. Older age at cessation and a diagnosis of ulcerative colitis were associated with a lower risk of relapse. This reflects existing data, with two previous studies showing a similar association with older age at diagnosis7 and at time of cessation.13 While the cohort was too small to analyse relapse by Crohn’s phenotype, as those with penetrating disease were excluded, the majority had an inflammatory phenotype.
In addition to optimising patient care by reducing unnecessary exposure to drug, regular panel review of biologic therapy can be cost-saving.16 In our experience, the costs of running the BRP are minimal, taking up four sessions (24 hours) for each specialist nurse, with two further sessions of preparation. In addition, three consultants each give six sessions per year (24 hours each), that is, a total cost of £19 600 over 5 years. While the costs of anti-TNF agents have reduced with the introduction of biosimilars, at the time of the study this was more than offset by cost savings from deprescription of anti-TNF therapy over the 5-year period.
As therapeutic drug monitoring became available to us mid-way through the study (from 2015), we have not analysed its role in this study. However, after 2015 it has formed part of the BRP discussion for most patients, allowing optimisation of therapy for those not in remission with inadequate drug levels and also identification of those in remission with no active drug in the system. This enabled a timely switch to other medical or surgical therapies, avoiding costs and potential morbidity from continued administration of ineffective medication.
Additional potential benefits from a review panel-style approach to patient care are in streamlining investigation and maximising efficiency of clinic appointments. These are difficult to quantify, and it is arguable that some patients would receive equivalent care without the BRP; however, at least 6 of 45 (13%) felt to be in clinical remission by their usual clinician were subsequently found to have active disease and excluded from biologic withdrawal on further investigation by the BRP. Additionally, all patients recommended for withdrawal subsequently had biologics withdrawn, suggesting that patients and clinicians had confidence in the panel decision. This may also reflect the impact of our robust local IBD nurse service with clear channels for direct patient communication and the ability to rapidly assess patients in the nurse-led clinic. This enabled efficient disease evaluation and outpatient retreatment when relapse has occurred, thus limiting potential harms of anti-TNF cessation. In future, with reducing costs of anti-TNF agents and concerns regarding the risks of long-term thiopurine usage,10 a similar approach could be used to identify individuals who would be suitable for maintenance anti-TNF or other biologic monotherapy without an additional immunomodulator.
There are some limitations to the data presented in this study, in particular the incomplete calprotectin data in patients at the beginning of the study period, with more comprehensive use of calprotectin testing prior to cessation more recently. This reflects the increasing use of this assay in our trust over this period. In addition, while all patients were clinically assessed prior to cessation, CDAI scores were not routinely calculated. Nevertheless, this reflects the reality of patient assessment in the clinic, and our findings demonstrate the real-world effectiveness of a focused review panel approach to important clinical decisions in the management of patients with IBD on biologic therapy.
In conclusion, our data demonstrate one of the lowest reported relapse rates in a series of elective anti-TNF cessation in IBD. We have shown how the use of a review panel, excluding those at high risk of relapse, can optimise selection of suitable patients. Proactive de-escalation of biologic agents to maintenance immunomodulator therapy in carefully selected individuals has the potential for significant cost savings. Further work to refine inclusion and exclusion criteria should make such an approach even more effective.
The authors would like to thank Dr Mathis Heydtmann, Dr Rizwana Hamid and Ms Fiona McLuskey for their contribution to developing and for participating in the BRP.
Contributors RS is the guarantor of the study. All authors (GN, SIS, RS, CL, AB, SL and LLC) participated in data collection and drafting the manuscript. The original idea for the study was devised by GN, AB and SIS, and further developed by RS. GN, SL, LLC, SIS and AB conducted the BRP intervention.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient consent for publication Not required.
Ethics approval As the review panel was introduced as part of a service development initiative and formed the standard of care for all patients, external ethical approval was not sought.
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement Raw anonymised data are available upon reasonable request from the corresponding author.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.