295endpoint of clinical remission at either treatment dose in comparison to placebo
(4.3% in the pooled etrolizumab treatment groups). Etrolizumab continued to dem-
onstrate safety and tolerability. No serious infections were reported in the etroli-
zumab treatment groups, and there was no difference in drug-related adverse effects
or adverse effects requiring discontinuation of therapy [ 12 ].
Several novel aspects of this trial make it a milestone in trial methodology and
indeed IBD therapeutics. Firstly, the percentage of patients on placebo achieving
remission at week 10 was zero. Placebo response rates in previously reported trials
have ranged between 5.4 and 14.9% [ 5 , 69 , 70 ]. The rigorous use of central reading
of endoscopic videos had never been implemented previously in trials of biological
therapies but has recently been demonstrated to be of vital importance in correct
patient enrollment [ 12 , 71 ]. The assessment of the primary endpoint at 10 weeks
may have contributed to the results and indeed may also have clinical implications
for the assessment of response for anti-integrin therapy and the lessons learned from
6-week assessments in the vedolizumab studies [ 43 , 45 ].
Pharmacodynamic studies of etrolizumab including the analysis of β7 occupancy
and expression on T- and B-lymphocyte subsets in peripheral blood and colonic tis-
sue, quantification of αE+ cells and gene expression of cytokines and adhesion mol-
ecules have possibly provided the first insights into additional predictive markers
for response to treatment [ 12 , 72 ]. Gene expression studies in colonic biopsies from
anti-TNF-naive UC patients treated with etrolizumab showed higher baseline
expression of T-cell-associated genes, including αE integrin and granzyme A mes-
senger RNA (GZMA) in clinical remitters and also higher rates of mucosal healing
in αEhigh and GZMAhigh patients [ 72 ]. The potential for predictive biomarkers has
exciting implications for further basic science research and indeed for personalised
medicine in the complex pathobiology of IBD. Etrolizumab is currently being eval-
uated in phase III studies in both CD and UC (Clinical trials.gov).
Alicaforsen (Anti-ICAM-1)
Alicaforsen (ISIS 2302) is an antisense oligonucleotide directed against human
intercellular adhesion molecule (ICAM)-1 and is expressed at low levels in endo-
thelial cells and leucocytes. Interest in ISIS 2302 as a potential therapeutic agent
stemmed from the observation that ICAM-1 is overexpressed in the inflamed intes-
tine of CD patients [ 73 ]. In a pilot study of 20 patients with active CD, 13 infusions
of different doses of ISIS 2302 or placebo over 26 days, ISIS 2302 demonstrated
superiority over placebo for corticosteroid-free remission [ 74 ]. A subsequent
double- blind, placebo-controlled trial did not show significant efficacy between 2-
and 4-week alicaforsen groups compared to placebo (20.2% and 21.2% vs. 18.8%)
[ 75 ]. Schreiber and colleagues conducted a dose-interval, multicentre, placebo-
controlled trial in 75 patients with steroid-refractory CD [ 76 ]. The primary endpoint
(steroid-free remission at week 14) was only achieved in 2 of 60 (3.3%) alicaforsen-
treated patients and none in the placebo-treated group. No further studies have been
performed using alicaforsen in CD.
16 Anti-integrin Agents in IBD: Efficacy and Risk of Complications