Anti-tumour necrosis factor-α treatment for perianal Crohn’s disease in Australia

Perianal Crohn’s disease (PCD) is a complication of Crohn’s disease (CD)1 that can significantly impair patients’ quality of life.2,3 Anti-tumour necrosis factor-alpha (anti-TNFα) treatment provides a significant advance in management of PCD-associated fistulae. The effectiveness of anti-TNFα therapy for alleviating PCD (including closing fistulae), first demonstrated by Present and colleagues4 in a double-blind randomised controlled trial (RCT), has been confirmed and extended by two other double-blind RCTs: the ACCENT II Trial (A Crohn’s Disease Clinical Trial Evaluating Infliximab in a New Long-term Treatment Regimen in Patients with Fistulizing Crohn’s Disease)5 and the CHARM Trial (Crohn’s Trial of the Fully Human Antibody Adalimumab for Remission Maintenance).6 These trials demonstrate a reduction in hospitalisations and surgical procedures as well as improved health-related quality of life with anti-TNFα therapy.5-8 They are the only medical therapies shown by double-blind RCTs to alleviate PCD-associated fistulae.

Currently, two anti-TNFα agents, infliximab and adalimumab, are listed on the Australian Pharmaceutical Benefits Scheme (PBS) for treating subgroups of CD patients with difficult active luminal disease, extensive small bowel disease, a short gut or a stoma (Box 1). Clinical criteria for PBS subsidy rely in large part on the Crohn’s Disease Activity Index (CDAI)9 as a measurement tool to assess eligibility. Other secondary measures of intestinal inflammation (raised platelet count, erythrocyte sedimentation rate, C-reactive protein level, stool lactoferrin or calprotectin level, and radiological testing) are considered for patients with small bowel disease, short gut syndrome, ileostomy or colostomy.

Unfortunately, the CDAI score, which was developed to measure luminal disease activity,10 does not reflect PCD activity. Patients primarily affected by a draining fistula may have a relatively low CDAI score, as this contributes only 20 points to the overall CDAI score.11 Hence, the PBS criteria potentially exclude patients with PCD from obtaining subsidised anti-TNFα medications, despite Level 1 evidence of clinical efficacy of these agents for PCD and the lack of adequate alternative therapies.

The aims of our study were to assess the prevalence of PCD among patients in four large Australian inflammatory bowel disease (IBD) centres and to determine the proportion of patients with clinically significant PCD who would qualify for subsidised anti-TNFα therapy under current PBS rules.

We conducted a cross-sectional cohort study of adults from four Australian IBD centres over the period January 2004 to May 2008, using an established AIBDA (Australian IBD Association) web-based database. AIBDA is one of three specialty interest associations within the Gastroenterology Society of Australia. Its members have an active clinical research focus on IBD. Data in the database are collected prospectively but voluntarily. Contributing centres were the Mater Health Services Adult Hospital, Brisbane, Queensland; the Royal Brisbane and Women’s Hospital, Brisbane, Queensland; Fremantle Hospital, Fremantle, Western Australia; and Flinders Medical Centre, Adelaide, South Australia. Each of the contributing centres was asked to provide site statistics on:

• The total number of IBD and CD patients;

• The total number of PCD patients;

• The underlying luminal disease distribution in PCD patients;

• The number of PCD patients for whom anti-TNFα therapy was clinically indicated; and

• The number of patients with PCD who met the current PBS prescribing criteria.

For the purposes of our study, PCD was defined as “active fistulising perianal, vulvovaginal and/or penile disease”. Patients with skin tags, haemorrhoids, fissures or isolated anorectal strictures were excluded. Luminal disease distribution was classified as ileal, colonic, ileocolonic or upper gastrointestinal, according to the Montreal disease classification.12 In addition, there was a group of patients in which disease was confined to the perianal or genital regions without intestinal involvement. The treating physicians were asked to retrospectively assess eligibility for PBS-subsidised anti-TNFα therapy according to the current criteria for the CD subgroups (Box 1). Anti-TNFα therapy was deemed to be clinically indicated, irrespective of current PBS criteria, for patients with moderate or severe faecal or mucopurulent discharge from fistulating PCD4 in spite of adequate conventional treatment (defined as at least 4 months’ therapy with an adequate dose of thiopurine or methotrexate, with or without antibiotics).4

Complete data were identified for 3589 adult patients from the participating IBD centres (Box 2). Of these, 1815 (51%) had CD, and of those with CD, 310 (17%) had PCD.

Among patients with PCD, the largest group was those with ileocolonic disease (143 [46%]), followed by colonic disease (84 [27%]) and ileal disease (71 [23%]). Twelve patients (4%) had PCD in the absence of luminal disease. None of the PCD patients had isolated upper gastrointestinal CD.

Anti-TNFα therapy was deemed clinically indicated (as defined in the Methods section) for 54% of patients (Box 3). Seventy per cent of these patients met current PBS criteria for subsidised medication (113 on the basis of active luminal disease, three on the basis of short gut, ileostomy or colostomy, and one on the basis of extensive small bowel disease). There was variation between sites with regard to the proportion of patients with PCD requiring anti-TNFα therapy (P < 0.001) and the proportion of patients eligible for PBS-subsidised treatment (P = 0.001).

Our study sampled a large population of IBD patients, which we estimate (on the basis of a 2007 report on CD13) to be 6% of all adult and paediatric patients with IBD in Australia. Patients with CD made up slightly over half of our study population. The prevalence of PCD (17%) was consistent with prevalences reported elsewhere.14,15 Nearly a third of PCD patients in our study who had a clinically indicated need for anti-TNFα therapy did not meet PBS criteria and therefore could not receive subsidisation for these high-cost drugs. This occurred despite Level 1 evidence demonstrating the efficacy of anti-TNFα therapy for fistulising CD and the lack of effective alternative therapies.4-8

Eligibility for PBS subsidy is strongly influenced by the CDAI score, but the score is a poor measure of quality of life. Inadequacy of the CDAI for scoring PCD severity may explain why many patients for whom anti-TNFα therapy is clinically indicated are ineligible for PBS subsidy. This was highlighted by a recent Australian study described during Australian Gastroenterology Week.16 The study found that in a cohort of patients with active PCD resulting in significant impairment to quality of life, the mean CDAI score was only 236, with individual scores rarely exceeding the threshold of 300 needed to comply with PBS criteria.

Several methods have been devised to quantify PCD severity. In our study, we employed a “fistula drainage assessment” of moderate or severe faecal or mucopurulent discharge to define clinical severity. This assessment tool, used in the first infliximab double-blind RCT,4 has the advantages of being simple, easy to use retrospectively (because the presence or absence of an actively discharging fistula is recorded in our database assessment field), and clinically relevant to anti-TNFα treatment. The more comprehensive assessment tool, the Perianal Disease Activity Index (PDAI),17 which scores five subcategories (discharge or drainage; pain; restriction of sexual activity; type and number of perianal lesions; and induration), is not fully clinically validated and is difficult to equate with meaningful clinical remission.10 Furthermore, anti-TNFα treatment can be contraindicated if there is induration (the fifth subcategory of the PDAI).

There was significant variability among patients in our study with regard to clinical indication for anti-TNFα therapy and eligibility for PBS-subsidised therapy (Box 3). While this is perhaps not surprising, given that this is a real-life post-marketing study based on information from a voluntarily maintained database, the causes of variability require some explanation. First and foremost, the large subjective component involving functional assessment of diarrhoea, pain and wellbeing contributing to the CDAI score may lead to significant variation or even manipulation of the CDAI score with regard to eligibility. Other factors that may contribute to variability include different referral patterns within Australian states and individual prescriber thresholds for failure of conventional therapy.

The total cost to the government of anti-TNFα therapy in the financial year 2007–2008 in Australia was $174 397 235,18 most of which was expended on rheumatoid arthritis. Expenditure on CD was $14 125 377. Based on prevalence data, it is estimated that the anti-TNFα expenditure on all PCD in 2007 was about $2 400 000. This cost should be compared with the cost of not providing PBS-subsidised anti-TNFα treatment for PCD. According to a recent Australian study, CD patients with fistulising disease incur a significantly higher annual health care utilisation cost (exclusive of medications) than CD patients without fistulae.19 This includes hospitalisation for local surgery (eg, drainage or repair of fistulae) and more radical surgery (eg, ileostomy). Importantly, from the patient perspective in this predominantly young population, anti-TNFα therapy can help to alleviate faecal incontinence and to reduce awkwardness around sexual activity. We believe that the cost of expanding the PBS criteria to include all patients with clinically significant PCD would be well justified, given the proven clinical benefit of anti-TNFα therapy in this group of CD patients.