Molecular Heterogeneity, Biomarker Discovery, and Targeted Therapy in Asthma - Genomic Biomarkers for Pharmaceutical Development - page 79

found that serum periostin, FeNO, and to a lesser extent, blood eosinophils increased across
the population with increasing levels of airway eosinophilia. Of these three biomarkers,
serum periostin predicted airway eosinophilia with greater sensitivity and specificity than
either FeNO or blood eosinophils and exhibited less intrapatient variability over time [46] ,
although as discussed above, the three measures reflect similar biological processes and
thus it was unsurprising that they were intercorrelated, if weakly, across the population.
4.3 BIOMARKER-GUIDED PATIENT SELECTION IN
THERAPEUTIC TRIALS
Several investigational therapies targeting type 2 inflammation in asthma have undergone
early clinical development. Detailed discussions and listings of these agents can be found
elsewhere [26-28] ; we will focus here on proof-of-concept studies that have used biomark-
ers to select patients most likely to demonstrate clinical benefit from agents targeting IL5 and
IL13, as agents targeting these cytokines have the largest amount of publicly available clinical
trial data.
4.3.1 Agents Targeting IL5
As discussed above, IL5 is an obligate hematopoietic factor for eosinophils, regulating
their differentiation, activation and survival. Three humanized monoclonal antibody thera-
pies directed at IL5 or its receptor have been evaluated in randomized, placebo-controlled
trials in asthma patients. The first agent to be described was mepolizumab (anti-IL5), which
failed to demonstrate clinical benefit in an unselected population of moderate asthma
patients incompletely controlled on ICS therapy, despite showing significant pharmacody-
namic effects on blood and sputum eosinophils [61] . Three possible explanations for the lack
of observed clinical benefit despite evidence for a biological effect in the study are:
1. Incomplete biological effect on eosinophils in bronchial mucosal tissue
2. Selection of a primary outcome measure - in this case, the asthma control questionnaire
(ACQ) [62] - that did not accurately reflect the relationship of the biological target to the
clinical presentation of asthma
3. Inadequate selection of those patients whose disease is likely to be associated with
eosinophil activity.
A non-significant trend toward reduction in asthma exacerbation rates (a secondary out-
come measure) was observed in the study, although there were not enough exacerbation
events during the period of observation to have sufficient statistical power to adequately
evaluate this outcome measure [61] . Two subsequent studies addressed the issues of out-
come and patient selection by enrolling only severe asthma patients with greater than 3%
sputum eosinophils who had experienced multiple exacerbations in the prior year [63,64] , as
past exacerbation history is a prognostic indicator of future exacerbations [65] . Each of these
studies demonstrated significant clinical benefit of IL5 blockade in terms of asthma exacerba-
tion reduction. Reslizumab, another anti-IL5, was investigated in moderate-severe patients