A U.K. team has identified a prognostic gene expression signature in CD8+ T cells that could be used to guide treatment decisions in patients with inflammatory bowel disease.1 The researchers are now planning a prospective clinical trial to verify the validity and utility of the signature.

IBD is an autoinflammatory condition that encompasses Crohn's disease (CD) and ulcerative colitis (UC), which have a related etiology but distinct clinical presentation. Treating UC and CD is complicated because clinical symptoms do not effectively predict the extent of disease progression.

Thus, doctors and patients are largely in the dark when weighing treatment options, which range from corticosteroids or azathioprine to anti-tumor necrosis factor-a (TNF-a) therapy to surgery.

Multiple genetic and immunologic markers have been linked to IBD, including mutations in caspase recruitment domain family member 15 (CARD15; NOD2) and the presence of certain antibodies.2 However, the prognostic utility of these markers is unclear.3

Mark Curran, senior director of immunology biomarkers at the Centocor R&D unit of Johnson & Johnson, said there is a need for better markers in IBD prognosis. "People have tried hard to look at clinical symptoms and signs, and it's failed because of the complex presentation of the disease. The objective has been to come up with a good biomarker; unfortunately the effect size of those looked at so far has been very small."

Ken Smith and his team at the University of Cambridge hypothesized that specific classes of immune cells might hold the key to a better prognostic marker for IBD patients.

Previous searches for prognostic signatures have included gene expression profiling of whole blood. "There were occasional signatures associated with disease activity and diagnosis, but they found surprisingly little. Nothing appeared of particular prognostic value," said Smith, who is head of the Department of Medicine at the university's School of Clinical Medicine.

One reason for the failure of prior efforts is variation in gene expression that occurs when multiple blood cell types are simultaneously profiled. To curb the variation, Smith used immunomagnetic microbeads to positively select and profile specific immune cell subtypes. His team isolated CD4+ and CD8+ T cells from 67 patients with CD or IBD and then performed a computationally unbiased whole-genome gene expression analysis on the samples.

The group found that gene expression signatures in CD8+ T cells could stratify patients into two distinct IBD groups, dubbed IBD1 and IBD2. Next, the team went back to determine if the two blinded groups had differences in disease progression.

Indeed, CD or UC patients classified in the IBD1 group had a significantly increased frequency of treatment escalation over a two-year follow-up period compared with the IBD2 group (p<0.003 and p<0.0002, respectively).

Results were published in The Journal of Clinical Investigation.

"These data suggest that the signature we have identified may represent the first method by which treatment strategies could be appropriately personalized at diagnosis," said Smith. "Such preemptive stratification would be expected to improve the therapeutic outcome in IBD patients destined to run a refractory course and avoid unnecessary immunosuppression in those with indolent disease."

"This work is very tantalizing; it is absolutely a useful paper for the stratification of patients," said Daniel Vitt, CSO of 4SC AG. "However, we've seen many efforts to use biomarkers and genetics to stratify patients."

He cited anti-Saccharomyces cerevisiae antibodies and variance in histocompatibility complex genes as markers for IBD that looked promising but failed to be predictive in the clinic.

4SC's vidofludimus is a small molecule inhibitor of dihydroorotate dehydrogenase (DHODH), IL-17A, and IL-17F that has completed Phase IIa testing in IBD.

Predicting treatment

Next year, Smith plans to start a trial in which the CD8+ gene signature will be used to assign treatment options to IBD patients at the School of Clinical Medicine and collaborating centers. The trial will compare the outcome of stratifying therapy according to the IBD1 and IBD2 subgroups with current treatment strategies, which adopt one-size-fits-all approaches.

The point of the trial will be to confirm that the signature enables more potent medical therapy (such as TNF-a inhibitors and azathioprine) to be commenced at diagnosis in patients who need them, as predicted by their CD8+ T cell gene signature, while avoiding unnecessary treatment for those who do not need it.

Current clinical guidelines recommend a stepwise approach to IBD treatment, in which patients generally start with systemic corticosteroids and only begin immunosuppressants such as azathioprine if their disease progresses. TNF-a inhibitors are usually reserved for patients whose disease progresses in spite of these second-line therapies.

"Anti-TNF-a therapy works best early, and patients benefit greatly with a high efficacy rate," but not every patient needs anti-TNF-a therapy to halt progression, said Stefan Schreiber, professor of medicine and gastroenterology at Kiel University. "If we can identify those patients who would progress, we could spare some patients who wouldn't need the therapy."

Smith added, "Most people who need anti-TNF-a therapy don't get it for a couple of years while other options are tried. Anti-TNF-a is less effective if it is delayed, but you don't know who will need it, and not everyone does."

A key unknown, said Curran, is whether there will be any correlation between the CD8+ signature and patient responses to TNF-a drugs.

"The biology underlying these signatures is not well understood. Until you test groups with a wide range of disease severity, you can't predict what effect a drug will have in each population. This is a point often missed when discussing the use of a predictive biomarker, but it must be taken into account in trial design," he said.

Curran is leading J&J's internal efforts to identify biomarkers for IBD and other autoimmune diseases. In March, the company published a gene expression analysis of patients experiencing differential responses to Remicade infliximab,4 and the company is seeking to understand why subsets of patients are refractory to the anti-TNF-a antibody.

J&J's late-stage clinical compounds for IBD include Simponi golimumab and Stelara ustekinumab. Simponi targets TNF-a and is in Phase III testing to treat UC. Stelara inhibits IL-12 and IL-23, and it is in Phase III trials to treat CD. The drugs are marketed for other autoimmune diseases.

Seeking insight

In addition to the trial plans, Smith's team is investigating whether the CD8+ signature can lead to the identification of new targets for autoimmune diseases.

There is reason to believe the signature extends beyond IBD. Last year, Smith published in Nature Medicine the identification of a CD8+ T cell gene signature that was useful in prognosis of antineutrophil cytoplasmic antibody-associated vasculitis and systemic lupus erythematosus (SLE).5 That signature is significantly similar (p<10-300) to the new signature identified in UC and CD patients.

"We didn't expect this signature would be the same in each of the diseases-in some ways there is no reason to assume it would be," Smith said. "Having said that, there are common signaling pathways at work in these diseases, similar SNPs are associated with them, and we know T cells are involved in these conditions."

Vitt agreed it is not surprising that SLE and IBD have similar signatures. "When we look in animal models for both indications with vidofludimus, we see similar effects. This work further underlies the overlap we see at the therapeutic level."

Vitt said the compound has the potential for decreased side effects compared with corticosteroids or anti-TNF-a therapies. The company is looking to partner vidofludimus for further late-stage clinical development. As a next step, the company is planning a Phase IIb trial.

Smith noted it is unclear why CD8+ T cells have the prognostic signature or whether it is related to disease progression. The IBD1 subgroup did provide some hints about mechanism and showed upregulation of T cell receptor (TCR) signaling, greater CD28 co-stimulation, and increased IL-2 and IL-7 signaling compared with the IBD2 subgroup.

Schreiber said methods such as gene expression profiling of immune cells do have the potential to help clarify the mechanisms underlying IBD. "This study says something deep; it demonstrates we can move away from animals, into patients, and measure gene expression reproducibly. This is important because there are no good animal models for these diseases."

Smith has filed a patent application covering the use of this method to predict autoimmune disease progression, and the work is available for licensing from Cambridge Enterprise Ltd., the technology licensing arm of the University of Cambridge.

Cain, C. SciBX 4(42); doi:10.1038/scibx.2011.1168
Published online Oct. 27, 2011


1.   Lee, J.C. et al. J. Clin. Invest.; published online Sept. 26, 2011; doi:10.1172/JCI59255
Contact: Kenneth G.C. Smith, University of Cambridge School of Clinical Medicine, Cambridge, U.K.
e-mail: kgcs2@cam.ac.uk

2.   Abraham, C. & Cho, J.H. N. Engl. J. Med. 361, 2066-2078 (2009)

3.   Vermeire, S. et al. Gut 55, 426-431 (2006)

4.   Toedter, G. et al. Am. J. Gastroenterol. 106, 1272-1280 (2011)

5.   McKinney, E.F. et al. Nat. Med. 16, 586-591 (2010)


      4SC AG (Xetra:VSC), Planegg-Martinsried, Germany

      Cambridge Enterprise Ltd., Cambridge, U.K.

      Johnson & Johnson (NYSE:JNJ), New Brunswick, N.J.

      Kiel University, Kiel, Germany

      University of Cambridge, Cambridge, U.K.