The medical management of IBD and how to avoid them

That oral aminosalicylates (active ingredient 5-aminosalicylic acid [5-ASA]) are not an effective treatment option for maintenance therapy in Crohn’s disease has been robustly demonstrated. Some have advocated for the use of 5-ASA to maintain remission in colonic Crohn’s disease, but there is no good quality evidence to support this, either from a dedicated colonic Crohn’s trial, or from either subgroup analysis or meta-analysis of existing maintenance studies. A recent Cochrane meta-analysis of studies using 5-ASA drugs in the maintenance of medically-induced remission went so far as to conclude that additional randomised control trials (RCTs) in this area may not be justified.² Indeed, it is noteworthy that even without such a meta-analysis, no single RCT has ever shown a statistically significant benefit of 5-ASA in maintenance therapy for Crohn’s disease. Although there has been some suggestion of benefit for the maintenance of remission of small-bowel Crohn’s disease after surgical resection, the effect sizes are extremely small.³ Expert consensus guidelines do not recommend the use of 5-ASA in Crohn’s disease.^4,5

Despite the evidence, 5-ASA drugs continue to be widely used for the management of around one third of patients with Crohn’s disease,⁶ a situation that is scarcely any better in large clinical trial centres.⁷ This may reflect a lack of cheap, nontoxic treatments for mild Crohn’s disease, as well as the legacy of conflicting information around 5-ASA formulations when they were first developed. Nonetheless, a lack of treatments and evolving understanding of efficacy does not equate to justification for inappropriate prescribing.

Longitudinal data may indicate that surgical rates for both Crohn’s disease and ulcerative colitis are declining, but surgery remains a very real possibility for all IBD patients.^8,9 Though definitive data are lacking regarding causality, advances in the medical management of IBD do at least correlate with the decline in surgery. These advances in medical management, together with concerns shared by patient and doctor alike regarding the irreversibility of bowel resection, may lead some to regard surgery as a last resort, to be delayed or avoided by any means possible, except in well-recognised situations of medical treatment failure (figure 1).

The LIR!C trial continues an important trend towards re-evaluating attitudes towards surgery, in particular with respect to the potential role of timely, early surgical intervention in a well-selected cohort. The LIR!C study recruited patients without a prior history of surgery or biologic therapy and who had limited, nonstricturing, nonpenetrating ileal disease that was refractory to thiopurine or steroids—in other words, a cohort that would, in many centres, be managed by escalation to biologic therapy. Indeed, the well-established nature of this medical treatment paradigm may explain why the study was somewhat slow to recruit to. Participants were randomly allocated to infliximab therapy or to laparoscopic ileal resection. Initial 1-year follow-up data from this pivotal study suggests that surgery is safe, efficacious, cost effective and associated with similar quality scores when compared with medical therapy.¹⁰ Longer-term outcome data and subgroup analysis from this study are keenly awaited.

The decision to consider surgical intervention should encompass a review of current disease burden and medical options, but also patient preference and psychosocial factors, as well as nutritional status (including any potential for further optimisation versus deterioration with delay). These factors must all be judged on an individual basis, but it is also important not to stall when a surgical approach is deemed necessary to avoid an adverse impact on postoperative recovery. If a patient is indeed an operative candidate, it is critical their case is highlighted early and discussed within a multidisciplinary meeting, since a successful outcome will require the expertise of surgeons and gastroenterologists working alongside radiologists, pathologists, IBD nurses, dieticians, stoma therapists and psychologists.

Figure 1 | Specific scenarios for when to initiate consultation with surgical colleagues.

In the words of G. Gavin Miller, “No attempt is made here to suggest that total colectomy is the final word on treatment for ulcerative colitis, nor even to insinuate that it is the right treatment for the majority of patients suffering from this disease. However, when a ship is foundering, a life boat, though neither pleasant nor agreeable, may be the only hope.”¹³
Total colectomy was pioneered by Miller in the late 1940s as a treatment for severe ulcerative colitis. Given the paucity of effective medical therapies at that time, it is perhaps unsurprising that surgical approaches became increasingly popular, despite the crudity of contemporary stoma appliances. Today’s patients have the benefit of huge advances in perioperative management, intraoperative techniques and stoma appliances. Many will also subsequently elect to undergo ileoanal pouch surgery, which is increasingly performed laparoscopically. Perhaps this surgical sophistication alongside historically rather limited choices for the medical management of moderate/severe disease led to the perception of surgery as the definitive cure for ulcerative colitis. Indeed, despite recent progress in medical therapy, rates of elective colectomy for ulcerative colitis have declined only gradually over the past two decades.⁹

Whilst for some patients the desire to be rid of symptoms will dominate any willingness to consider further medical treatment, it is still important to ensure they receive appropriate counselling about the post-surgical state. When appropriate, ‘normal’ pouch function, awareness of pouchitis and the need for pouch revision surgery should be clearly explained. Due consideration and discussion should also be given to the effect of surgery on sexual activity and fecundity—both in terms of the ability to form and establish sexual relationships and on subsequent fertility (see mistake 7).

Since the procedures involved are, to a greater or lesser extent, scarring and irreversible, patients must be carefully selected, appropriately counselled and psychologically prepared. For most, it is important to know that they, and their team, have at least considered, if not tried, all alternatives. This reassurance will be felt most strongly in centres where close liaison between medical and surgical teams is evident to patients and where there are opportunities for early discussions with the surgical team to allow time for consideration, rather than a one-way referral system after medical ‘failure’ (see mistake 2).

The ability to discuss experiences and expectations with a stoma nurse and patients who have undergone colectomy, either in person or via internet support groups, may also be useful for patients contemplating surgery. Medical treatments that can reduce the need for, or even just delay, surgery, may be especially valued by some patients, especially those who are younger or relatively newly diagnosed who have not had sufficient time to absorb the impact of their disease. Sensitive, timely but frank discussions about the possibility of surgery, sometimes even when planning the next medical therapy, will afford patients the opportunity to consider and prepare for these options.

The use of symptom-based scoring indices, such as the Crohn’s disease activity index (CDAI) or the Harvey–Bradshaw Index (HBI), as a marker of disease severity has been shown to result in inaccurate assessment of the true inflammatory burden of activity.^14,15 Nevertheless, national prescribing guidelines have frequently mandated the use of such indices to determine funding for biologic therapy.

Clinical trial endpoints have now shifted to reflect the importance of using robust biochemical, endoscopic and histological parameters to assess treatment response;¹⁶ it is now clear that patients who achieve a treatment response based on these parameters have improved long-term outcomes.^17–19 At the same time, there is increasing evidence from randomised controlled trials RCTs that, in certain situations, timely patient assessment with appropriate treatment escalation driven by clinical, endoscopic and biochemical parameters can result in improved outcomes.^20–22 Whether early treatment escalation in patients with Crohn’s disease who fail to achieve an endoscopic response can impact on patient-relevant medium/long-term outcomes is an important question and the subject of the ongoing REACT2 trial (ClinicalTrials.gov identifier NCT01698307; results not yet available). Furthermore, it is unclear which outcomes are the most relevant and meaningful in routine clinical practice.²³

Lack of absolute clarity on these points should not prevent a shift towards paying greater attention to variables that can be pragmatically monitored, such as normalisation of CRP levels and reduction of faecal calprotectin levels, as well as due consideration of endoscopic, histological and radiological information, when available. The CALM study has demonstrated that a composite approach, basing treatment escalation on biochemical parameters alongside clinical indices, led to significantly improved endoscopic responses in patients with Crohn’s disease at 48 weeks when compared with treatment based upon clinical indices alone.²²

As we currently lack validated biomarkers for the reliable identification of patients at high risk of disabling disease, contemporary treatment algorithms continue to adopt a ‘step-up’ approach,^4,5 albeit with increasing emphasis on early identification of treatment nonresponders using objective measures (as discussed in mistake 4). In the context of the step-up model, it is widely recognised that timely escalation of therapies is essential to avoid patients languishing too long on inappropriate drugs that are failing to control their condition.

Prior to treatment escalation, it is vital to ensure first that the existing treatment has been used optimally and second that the patient’s current symptoms are reflective of active IBD. The first caveat requires assessment of compliance, and appropriate dose optimisation and therapeutic drug monitoring when available, all of which may help maximise the value of existing therapies. The second point requires due consideration of all alternative diagnoses that may coexist in an IBD patient, including functional symptoms, which are present in around one-third of IBD patients.²⁴ Other important causes of worsening gastrointestinal symptoms in a patient with IBD include infections (e.g. Clostridium difficile and CMV), NSAID use, small-bowel bacterial overgrowth and bile salt malabsorption (the latter two are particularly common in patients who have undergone a prior ileocaecal resection).This approach mandates appropriate combined clinical, endoscopic, biochemical, microbiological and radiological assessment prior to any change in treatment. The increased availability of faecal calprotectin testing has undoubtedly helped in this regard (see mistake 4) and may be predictive of an IBD flare event prior to clinical symptoms appearing.²⁵

How tests perform differs in the IBD population and between subpopulations with variable extents of colonic disease, which has led to a potentially confusing range of proposed cut-off values, mostly from post-hoc and retrospective analyses. We believe that the best use of faecal calprotectin in this context may be incorporating serial assessment during periods of remission to establish a baseline from which any significant departure can be investigated through prompt further assessment.

Corticosteroids are effective for the induction of remission in both Crohn’s disease and ulcerative colitis, but are not an effective maintenance therapy in either the conventional form or as budesonide.^26,27 In addition, corticosteroids have well-documented side effects, such as an increased risk of infection, avascular bone necrosis, mood disturbance, hypothalamic-pituitary-adrenal axis suppression, osteoporosis, Cushingoid appearance and hypertension.²⁸ When compared with immunomodulators and biologic therapies, prolonged use of corticosteroids remains the single greatest risk factor for increased morbidity and mortality in IBD patients.²⁹ Despite advances in therapeutics, the relative risk of steroid exposure for IBD patients in their first 5 years after diagnosis from 1994–2008 remained static at ~50%.³⁰

The achievement of ‘steroid-free remission’ is recognised as a treatment target by professional societies and patient bodies.^31,32 European Crohn’s and Colitis Organisation (ECCO) guidelines stress the need for avoidance of prolonged or recurrent corticosteroid courses and suggest that corticosteroid-dependant patients or those receiving more than one course of steroids in a year should be offered treatment escalation.^5,33 Nevertheless, rates of corticosteroid dependency and recurrent corticosteroid prescribing remain high and have changed little over time.³⁴ In a study of corticosteroid dependency or excess in a cohort of 1,176 unselected outpatients attending IBD clinics across the UK, we found rates of steroid dependency or excess of 14.9%; expert review of charts from these patients showed that measures taken to avoid excess prescribing were suboptimal or inadequate in almost half of the cases, resulting in a rate of potentially avoidable steroid excess in this cohort of 7.1%.³⁵ Particular problems identified included patients taking self-administered courses of steroids which the secondary care team were unaware of, as well as patients receiving steroid prescriptions that were either inappropriate or too short to be effective. Such prescriptions were significantly more likely to be initiated in primary care, where there was a lack of appropriate communication with the secondary care team and of identification of the need for treatment escalation.

Finally, it is important to remember that corticosteroids have no role in the management of fistulae in patients with perianal Crohn’s disease, for which their use is associated with an increased risk of abscess formation and sepsis.³⁶ Taken together, this information suggests an increased need to identify and monitor steroid usage in patients as part of a well-configured and responsive IBD service.

A surprisingly common misconception is that patients with a chronic disease are not sexually active, do not rank sexual activity as important and will not become pregnant. In fact, sexual activity in patients with IBD forms part of validated, disease-specific quality of life (QOL) scores, developed based on patient consultation exercises.^37,38 However, patients with active disease do indeed experience significant body-image dissatisfaction and problems with sexual satisfaction.³⁹

Fertility in ulcerative colitis patients who have not undergone surgery appears normal and women with Crohn’s disease have lower fertility during disease flares only.⁴⁰ Nevertheless, fecundity in IBD patients, as measured by actual rates of childbirth, does appear to be significantly lower than in the general population,⁴¹ leading to the concept of ‘voluntary childlessness’. This is probably a misnomer, as a direct, inverse correlation between rates of voluntary childlessness in IBD patients and levels of knowledge relating to some basic facts around IBD in pregnancy has been shown.⁴² In general, patient scores on such basic knowledge tests were low, falling behind results achieved by nonspecialist ward nurses,^42,43 and revealed widely held misconceptions regarding, for example, the safety of IBD therapies during pregnancy.⁴⁴

Linking these observations together, we suggest:

• Evaluation of the impact of IBD needs to extend beyond simple enquiries about stool frequency, rectal bleeding and abdominal pain to include other areas of a patient’s life, including sexual functioning. These QOL metrics are increasingly incorporated into clinical trial data and scrutiny needs to be given to the effects of any potential new treatment across all these domains.
• The counselling of both male and female IBD patients of reproductive age regarding the impact of IBD and associated therapies on fertility should begin in the preconception phase. Waiting for a patient or partner to become pregnant before starting this discussion risks leaving patients to make important life decisions based on limited, often erroneous information.
• For female patients who are considering pregnancy, it is important to reinforce basic prepregnancy advice, including a review and discussion of safety around all current medications, to avoid patient/nonspecialist-initiated drug discontinuation, risking IBD flares during pregnancy with associated adverse pregnancy outcomes.

A more detailed discussion of the issues surrounding IBD and reproduction, can be found in another article from this series, “Mistakes in inflammatory bowel disease and reproduction and how to avoid them” by Kanis and van der Woude.⁴⁵

The development of perianal complications of Crohn’s disease portends a worse prognosis with regard to surgical resection and failure to respond to anti-TNF therapy.^46,47 The burden of disease is significant, with approximately 35% of all patients with Crohn’s disease experiencing one fistula episode during their disease course, of which 54% are perianal.⁴⁶ Combined surgical and medical management most likely achieves better outcomes for patients than either treatment alone.¹¹ However, the problem remains that, despite the evidence, many patients face significant delays in induction of biologic therapy after initial surgical control of sepsis, which may result in inadequate medical therapy and delayed healing. In a 2016 audit of practice at three large UK teaching hospitals, the median overall delay from first presentation to anti-TNF treatment was 204 days, including substantial delays after the first surgical consultation (Nicola Fearnhead, personal communication).⁴⁸ Undoubtedly several factors contribute to the delay, including involvement of nonspecialist teams in initial care, but any delays in anti-TNF therapy due to concerns regarding undrained sepsis can be allayed by ongoing communication with surgical colleagues and appropriate coverage with antibiotics.⁴⁹

The risk of dysplasia developing in patients with longstanding colitis has led to widespread recognition of the need for ongoing surveillance. In such patients, ECCO guidelines stress the need for thorough and complete endoscopic surveillance of the mucosa either by chromoendoscopy or white-light endoscopy, starting around 8 years after symptom onset.⁵⁰ These same guidelines suggest that in patients with quiescent colitis and no histologic activity of disease on the initial screening colonoscopy, and assuming no significant family history of colorectal cancer or a diagnosis of primary sclerosing cholangitis (PSC), surveillance colonoscopy should be offered in 5 years. By contrast, other guidelines, including those from the US, suggest much more frequent surveillance, up to annually, even in low-risk patients such as those described above.⁵¹
The latest data demonstrate that it is the inflammatory burden over time, rather than activity at any one timepoint, that determines the need for regular surveillance.⁵² Other important factors to consider when assessing risk include disease extent, duration of disease, PSC, family history of sporadic colorectal cancer, dysplasia and severity of endoscopic and histologic inflammation. This personalisation of risk forms the basis of the approach suggested by ECCO (Table 2), amongst others, and can help reduce inappropriate and unnecessary endoscopy in those at lower risk.

The overall incidence of venous thromboembolism (VTE) is more than twofold higher in patients with IBD than in the general population, and does not differ between ulcerative colitis and Crohn’s disease.⁵³ VTE represents an important and preventable cause of morbidity and mortality in patients with IBD.⁵⁴

The risk of VTE is associated with underlying disease activity and is greatly increased during periods of hospitalisation,⁵⁵ leading to a focus on prophylactic anticoagulation with, for example, low-molecular-weight heparin (LMWH) in hospital inpatients.^56,57 This treatment should still be given in the context of disease-related gastrointestinal bleeding, except when the bleeding is severe. Although most VTE occur in the outpatient setting in those with risk factors such as recent hospitalisation,⁵⁸ appropriate anticoagulation during admission appears to reduce the risk of subsequent outpatient VTE.⁵⁹ There is no evidence to support routine thromboprophylaxis in outpatients with active disease, although prophylaxis should be considered during flares of active disease in those with risk factors such as a previous episode of VTE.⁵⁷

As shown by several large population based studies, the increased risk of VTE is also paralleled by a modest increase in the risk of diseases associated with arterial thromboembolism, including myocardial infarction and stroke.⁶⁰ There is limited evidence that control of underlying disease activity can reduce these risks, and patients should be counselled on the importance of modifiable risk factors for cardiovascular disease, including cigarette smoking.⁵⁶