The mainstays of treatment for Crohn’s disease are corticosteroid therapy and therapeutic diets in the acute phase. Patients who relapse frequently require immune suppression (usually with azathioprine, 6-mercaptopurine or methotrexate) or anti-tumour necrosis factor (TNF) antibody treatments such as infliximab. The majority of patients can be maintained in remission with the use of such treatments, but a significant minority cannot, either being primary non-responders or losing response after a period on therapy. The latter is a particular problem with anti-TNF treatments, and is suggested to relate to the development of a host immune response targeting the therapeutic antibodies.1 In such situations resectional surgery is often the only option, with the most recent data suggesting that this is still required in up to 80% of Crohn’s disease patients over the long term.2

One of the key goals of the recently deployed techniques of genome-wide association scanning is to provide new understanding of the biological processes that lead to complex disease. In this regard it has proven dramatically successful, and nowhere more so than in Crohn’s disease. Over the past 2 years more than 30 confirmed Crohn’s disease susceptibility genes and loci have been identified. Among the themes to have emerged is the key importance of interleukin 23 and the Th17 pathway, and of innate immune pathways, particularly in how they sense and respond to bacteria.3^–6 With regards to the latter, two major hits have been observed in the autophagy genes ATG16L1 and IRGM. These have highlighted the importance of dysregulated handling and sequestration of intracellular bacteria in Crohn’s disease pathogenesis, and echo the earlier finding of association between Crohn’s disease and variants in nucleotide binding and oligomerisation domain 2 (NOD2), an intracellular receptor for bacterial motifs.7

The imperfections of the current therapeutic armamentarium for Crohn’s disease provide continuing impetus to the search for new, effective and safe treatments. It is hoped that insights regarding biological mechanisms of disease gleaned from recent genetic advances will aid this process, ideally yielding “small molecule” therapies which selectively target one or more of the key pathogenic pathways. In this case report we describe a novel treatment used in a patient with Crohn’s disease who had become refractory to conventional medical therapy.

CASE REPORT

The patient, who worked in the pharmaceutical industry and had a background in biomedical science, presented to her local hospital in 1996 at the age of 25 with diarrhoea, bloating and passage of mucus per rectum (PR). Colonoscopy showed segmental colonic and ileal ulceration, with histology demonstrating granulomatous inflammation consistent with Crohn’s disease. She initially responded to exclusion diet and mesalazine, then developed an ischio-rectal abscess for which she received azathioprine and methotrexate. By 2005 she had developed a recto-vaginal fistula and started treatment with 8-weekly infusions of infliximab (5 mg/kg) with continuing azathioprine at 50 mg/day. Her disease remained well controlled on this regime until she relapsed in January 2007 with further diarrhoea and abdominal pain. In May 2007 her Harvey–Bradshaw index was calculated to be 13 and endoscopy showed severe pleomorphic ulceration throughout the colon, with ulcers up to 30 mm in diameter (fig 1A). She responded initially to a course of corticosteroid therapy and increased frequency of infliximab infusions to 6-weekly. However, by August 2007 she was again opening her bowels up to 10 times/day, with recurrence of symptoms from her recto-vaginal fistula and development of a further perianal abscess requiring surgical drainage.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 Macroscopic appearances at colonoscopy before (A) and after 16 weeks treatment with sirolimus (B). In each case, colonoscopy was performed 6 weeks after infusion with infliximab.

Treatment options considered at this point were alternative anti-TNF antibody therapies, novel immune suppressants, or surgery (defunctioning ileostomy or pan-proctocolectomy). In August 2007 she was started on sirolimus, initially at a dose of 2 mg/day, and remained on infliximab at 6-weekly infusions (5 mg/kg). The dose of rapamycin was adjusted until serum trough levels were 5 ng/ml and this was established on a dose of 4 mg/day. Her symptoms improved markedly and the Harvey–Bradshaw index was calculated to be 4 after 8 weeks of adjuvant treatment with sirolimus. She was having her bowels open 2–3 times per day with neither abdominal nor perianal pain, and was no longer experiencing symptoms from her recto-vaginal fistula. Furthermore, after 12 weeks on sirolimus we were able to increase the interval of infliximab infusions to 8 weeks.

She underwent repeat colonoscopy in December 2007: this was 6 weeks after her most recent infliximab infusion, the same as per the previous colonoscopy. There was a marked improvement in endoscopic appearance (fig 1B). Serum markers of inflammation had also fallen, with normalisation of her CRP, albumin and platelet count (table 1). By the end of January 2008, after 6 months remission, she expressed a wish to become pregnant, so her sirolimus was stopped. To date her infliximab infusions have been maintained at 8-weekly without further relapse.

DISCUSSION

This is the first case report of the use of sirolimus to treat Crohn’s disease. In a patient with severe refractory disease who was losing response to infliximab it appeared to induce symptomatic remission and normalisation of objective markers of inflammation. Sirolimus (rapamycin, Rapamune) is an immunosuppressant that acts by forming a complex with FK-binding protein-12, which in turn binds to mammalian target of rapamycin (mTOR) kinase and thereby inhibits mTOR action. mTOR inhibition prevents cell cycle progression from G1 to S phase and thereby prevents T cell proliferation. This differs from the mechanism of other immunosuppressants such as tacrolimus and cyclosporin, which inhibit calcinurin to block transcription of interleukin 2 (IL2), and thereby T cell proliferation. Sirolimus is currently recommended for use in renal transplantation to reduce or prevent graft rejection by the host.8 Among its known and predictable side-effects, which include increased risk of infections, mouth ulceration, arthralgia, dyslipidaemia and hypertension, is an inhibition of fibrogenesis, leading to poor healing of surgical wounds.9 This is potentially advantageous given the tendency for the formation of fibrotic stricture in Crohn’s disease. There have been case reports of upper gastrointestinal and ileal ulceration but only in the context of multiple other drugs in medically complex situations post-transplant.10 11 Another important attribute is the fact that rapamycin is also used experimentally to induce autophagy, again via its inhibition of mTOR.12 In the light of recent genetic findings (see below) this further adds to its attractiveness as a candidate drug for the treatment of patients with Crohn’s disease.

The importance of the autophagy pathway in the pathogenesis of Crohn’s disease has been flagged by the results of three recent genome-wide association scans in Crohn’s disease. These have provided a major advance in our understanding of the genetic susceptibility to Crohn’s disease; in particular highlighting variants in two genes critical to autophagy (ATG16L1 and IRGM) which show confirmed association with the risk of Crohn’s disease.3^–6 Autophagy is a cellular process whereby obsolete organelles are encapsulated in a double membrane and degraded by fusion with a lysosome. Furthermore, autophagy has been implicated as an important innate immune mechanism for the clearance of intracellular organisms such as Salmonella typhi, Listeri monocytogenes and Mycobacteria tuberculosis.13 Given these observations it is intriguing to speculate that defective autophagy might underlie the finding of non-pathogenic luminal bacteria, particularly adherent-invasive Escherichia coli, persisting within the tissue macrophages and intestinal epithelial cells of individuals with Crohn’s disease.14 As with NOD2, these new findings provide a further link between genetically determined variation in innate immune response to bacteria and susceptibility to Crohn’s disease.

It is important to emphasise that the precise mechanism by which the variants in ATG16L1 and IRGM predispose to Crohn’s disease is currently unknown. One hypothesis is that they might adversely impact autophagy and result in failure to clear intracellular bacteria, leading to recruitment of the adaptive immune response targeting the intestinal epithelium with subsequent inflammation. However, other possibilities exist, including adversely affecting processing of peptide prior to major histocompatibility complex (MHC) class II binding on antigen-presenting cells.15

In our patient there was a clear improvement in Crohn’s disease symptoms after 8 weeks treatment with sirolimus and this was sustained until the treatment was stopped at 6 months. The drug was well tolerated, with minor sleep disturbance the only side-effect. In view of the significant degree of immunosuppression with combined sirolimus and infliximab we were vigilant for opportunistic infection, and monitored the patient weekly for the first 3 months and 2-weekly thereafter. Whether concomitant therapy with infliximab augmented response to the sirolimus in this instance is unknown, but it was clear that prior to the introduction of sirolimus the infliximab was losing efficacy, which was evidently recaptured following introduction of the oral immunosuppressant. Intriguingly, this response appears to have been sustained over the 3 months since the sirolimus was stopped.

Given its attractive attributes as a candidate therapy for refractory Crohn’s disease there is now a need to properly evaluate the role of sirolimus in Crohn’s disease in a formal clinical trial.