Vol.:(0123456789) 1 3
Lung (2019) 197:107–109
https://doi.org/10.1007/s00408-018-0185-8
LETTER TO THE EDITOR
Iftikhar and Colleagues Reply: Methodology Clarifed
Imran H. Iftikhar
1
· Mathew Schimmel
1
· William Bender
1
· Colin Swenson
1
· David Amrol
2
Received: 21 November 2018 / Accepted: 26 November 2018 / Published online: 5 December 2018
© Springer Science+Business Media, LLC, part of Springer Nature 2018
We thank Gunsoy and colleagues for their comments on our
paper [1] and appreciate the opportunity to further clarify
our methods and results. The defnitions of asthma exac-
erbations difered in diferent studies and was provided
in each article mostly in their ‘methods’ sections. Yes, it
is true, that we ‘did not extract treatment efects’ directly
from the source articles. We simply used the asthma exac-
erbation data reported in each study (see Table for further
clarifcation) in our meta-analysis. We never reported that
some biologics increased asthma exacerbations as Gunsoy
et al. imply. It is true that ‘patient-years’ were not provided
in each study, which is why we specifed in the methods
section, that we calculated the patient-years based on the
data on follow-up provided in each study. However, it is not
true, that exacerbation rates were not reported ‘within most
studies’ (see Table). For Chupp et al. [2], Bel et al. [3], and
Castro et al. [4] studies, the asthma exacerbation ‘counts’/
rates/percentages (defnitions and reporting formats of which
varied) were reported in the respective papers (see Table).
Our intent for using the approach of calculating ‘person
years’ was to present the data in a uniform format across
a range of diferent studies that difered in their follow-up
periods. However, it is understandable how this may have
misled some of the readers. We take this opportunity to pre-
sent the results in the conventional format using just the
counts of asthma exacerbations from individual papers (this
time, however, excluding Nair [5], Pavord [6], and Hanania
[7] studies to avoid any further confusions on calculation of
asthma exacerbation counts), and simply calculating odds
ratios. The results (see Table) confrm that dupilumab and
reslizumab were least associated with asthma exacerbations,
and show similar results for mepolizumab. However, these
results just like the results of the analysis reported in our
original paper have important limitations: (1) only stud-
ies from which these data were extractable in a common
format were included and others were excluded, (2) since
these results are not a ‘network’ meta-analysis of asthma
exacerbations, they do not provide ‘indirect estimates’ of
‘between-drug’ comparisons, and (3) since the time when
we analyzed the results, two new much awaited studies [8,
9] on dupilumab have been published. With regards to the
latter, an updated network meta-analysis is necessary at this
time, since in ours, we only included two of the dupilumab
studies.
Study ID Drug
events
(%)
Placebo
events
(%)
‘Asthma exacerbation’
in source articles
Benralizumab studies
Nowalk 2014 43 47 Subjects with ≥ 1
asthma exacerbation,
n (%)—week 24
Table 2 Pg. 18
Park 2016 9 8 Reported as “serious
adverse event”
Page 141 text
Bleecker 2016 13 19 “Worsening asthma”
(during “the treatment
period”)
Pg 2124 3rd para
FitzGerald 2016 4 5 “Worsening asthma”
(during “on-treatment
period”)
Pg 2138, para 3
Ferguson 2017 1 2 AE (“worsening
asthma” in > 3% pf
patients)
Pg 573, 2nd para
Pooled odds ratio: 0.80 (0.54–1.19), p = 0.28
Reslizumab studies
Castro 2011 8 19 “Clinical asthma exac-
erbation”
Pg 1129, 2nd para
* Imran H. Iftikhar
imran.hasan.iftikhar@emory.edu
1
Division of Pulmonary, Allergy, Critical Care & Sleep
Medicine, Emory University School of Medicine, 613
Michael St, NE, Atlanta, GA, USA
2
Division of Allergy and Immunology, University of South
Carolina School of Medicine, Columbia, SC, USA