Search in the Abstract Database

Search Abstracts 2011

* = Presenting author

P069. Identification of subgroups of adalimumab-treated patients with Crohn's disease who experience high rates of deep remission

J. Colombel1, W.J. Sandborn2, E. Louis3, R. Panaccione4, M. Yang5, J. Chao5, P.M. Mulani5

1Centre Hospitalier Universitaire de Lille, Lille, France; 2University of California, San Diego, La Jolla, CA, United States; 3University of Liège, Liège, Belgium; 4University of Calgary, Calgary, AB, Canada; 5Abbott Laboratories, Abbott Park, IL, United States

Aim: Adalimumab is efficacious in inducing and maintain deep remission, defined as clinical remission (CDAI < 150) and mucosal healing (absence of mucosal ulceration), in EXTEND. We identified patient subgroups (hyper-responders) for whom adalimumab treatment provided a high probability of achieving deep remission.

Materials and Methods: EXTEND was a randomized, placebo-controlled study of patients with moderate to severe ileocolonic Crohn's disease (CDAI 220–450). Patients received open-label adalimumab 160-/80-mg induction therapy at Weeks 0/2 and were randomized at Week 4 to maintenance therapy with adalimumab 40 mg every other week or placebo. Hyper-responders were identified through a novel method that generated an incremental benefit index score (IBIS) for each patient. IBIS was generated as the differences in predicted values of deep remission rates through the regression models from adalimumab and placebo arms. The association between IBIS and the probability of achieving deep remission and the 95% confidence intervals were evaluated using non-parametric methods. Hyper-responders among patients using adalimumab were identified as patients with an IBIS threshold of more than a certain value, which corresponded to a ≥34% probability of achieving deep remission at Week 12. This rate is double the average deep remission rate (17%) of all adalimumab-treated patients in EXTEND. Validation of the IBIS model based on Week-52 deep remission data was conducted using ROC curves and C statistics.

Results: Deep remission hyper-responders at Week 12 (N = 7) were identified as patients with an IBIS threshold ≥0.32, which corresponds to a 34% probability of achieving deep remission. The hyper-responders tended to be younger and with shorter disease duration than non-hyper-responders (N = 57); hyper-responders were also more likely to have greater baseline IBDQ scores and to be naïve to anti-TNF therapy (table). ROC curves indicated excellent predictive value of IBIS with a C statistic of 0.92.

Week-12 Hyper-Responders Receiving Adalimumab Treatment
Characteristic IBIS ≥ 0.32 (N = 7)IBIS < 0.32 (N = 57)P-Value
Mean disease duration, yrs311<0.0001
Mean age, yrs25390.002
Prior anti-TNF therapy use, % of patients14510.07
Mean baseline IBDQ score1251080.09
Deep remission rate, % of patients7111<0.001

Other variables (eg, prior steroid use, prior IMM use, and baseline CRP) also contributed to the composition of IBIS scores, though only those with p < 0.1 were presented in the table.

Conclusion: Using the novel method of IBIS, we were able to identify a subgroup of the EXTEND population who were deep remission hyper-responders at Week 12 if treated with adalimumab maintenance therapy. These patients tended to be younger with shorter disease duration, greater IBDQ scores, and no history of anti-TNF therapy.