The Editor
BMC Pulmonary Medicine
Dear Sir,
We thank the Journal of BMC Pulmonary Medicine for accepting our work and providing us with further
opportunity to improve our paper. We have made the final revisions to the proof as suggested by the
editors.
We have addressed all reviewer comments and a point-by-point response is given below. The revised
manuscript and this cover letter have been uploaded. All changes have been highlighted in bold, blue text.
Yours sincerely
Lakmali Amarasiri
Corresponding author
Response to reviewer 1
Comment 1
(a)
Concerning the pH monitoring with dual sensors the data obtained on the proximal sensor are
lacking.
While pathological data at this level may suggest the phenomenon of microaspiration, but this is
another mechanism of the reflux induced asthma.
(b)
At the level of the distal sensor for the distinction of normal or abnormal reflux events ,in special
in asthmatic patients, other criteria than the DeMeester score alone is essential to explore. An
increase in the number of reflux events at the level of the distal esophagus is detectable in the
reflux induced asthma, while the total score remains normal.( see in: Roka R, Rosztoczy A,
Izbeki F,Taybani Z, Kiss I,Lonovics J, Wittmann T.: Prevalence of respiratory symptoms and
diseases associated with gastroesophageal reflux disease. Digestion, 71:92-96,2005.).By this
mean the selection of patients remains doubtful, especially in case of 10 patients who had normal
pH values but positive GERD symptom score. If all the DeMeester criteria were really normal in
these patints,the analysis of a correlation between symptoms and physiological reflux events is
mandatory to establish or rule out a hypersensitive esophagus.
Response
(a) The proximal and distal sensor data have been included as table 3 (see below) and included in
the results (see page 8 of revised manuscript).
Table 3 - 24 hour pH monitoring data of subjects
SPRP
SPRN
SNRP
SNRN
Proximal sensor (15cm above LES)
Total No of reflux episodes
No of reflux episodes> 5min
Longest reflux episode
Total % of time pH<4
% of time pH<4 in upright position
% of time pH<4 in supine position
Demeester score
100.0±23.2
5.8±1.5
71.5±14.8
9.8±3.4
7.0±2.8
12.1±5.3
45.2±12.1
14.7±7.2
0.6±0.3
13.0±12.8
0.2±0.05
0.2±0.06
0.1±0.0
3.7±2.3
87.7±25.3
3.7±1.2
45.7±17.3
4.9±1.3
8.1±2.1
5.3±2.8
31.1±7.3
12.1±2.8
1.7±0.6
2.9±1.4
0.3±0.07
0.3±0.09
0.3±0.2
2.8±0.4
Distal sensor (5cm above LES)
Total No of reflux episodes
No of reflux episodes> 5min
Longest reflux episode
Total % of time pH<4
% of time pH<4 in upright position
% of time pH<4 in supine position
Demeester score
103.2±20
6.4±0.9
58.8±15.2
14.6±4.5
9.7±3.3
20.2±6.2
57.0±12.2
31.5±7.3
10.8±8.1
4.8±1.7
0.7±0.2
1.5±0.7
0.7±0.3
6.3±1.3
64.9±13.8
4.3±0.9
51.2±15.4
8.8±2.0
7.5±2.4
5.9±2.0
31.4±5.4
22.7±7.2
1.6±0.8
2.9±1.1
0.5±0.2
1.1±0.7
0.5±0.2
4.3±1.2
* All values as mean ± SE
(b) We employed the criteria of abnormally increased distal sensor Demeester score as the cut-off
to determine reflux positive state. Based on this we had 10 subjects per category.
As suggested, when the 10 patients who had normal distal pH values but positive GERD
symptom score were reexamined using total number of reflux events as the criteria to
determine abnormal reflux, we discovered that 4/10 of them had abnormal proximal reflux. We
admit that this was a limitation of the study.
We have included a statement acknowledging this as follows (see page 11 of revised
manuscript).
“We employed the criteria of abnormally increased distal sensor Demeester score as the cut-off
to determine reflux positive state. Based on this we had 10 subjects per category. A previous
study demonstrated that an increase in the number of reflux events at the level of the distal
esophagus is detectable in the reflux induced asthma, while the total score remains normal.
(Roka R, Rosztoczy A, Izbeki F,Taybani Z, Kiss I,Lonovics J, Wittmann T.: Prevalence of
respiratory symptoms and diseases associated with gastroesophageal reflux disease. Digestion,
71:92-96,2005.) Failure to consider this is a limitation of the study.”
Comment 2
The authors discussed the role of mucosal damage in the development of the esophago-bronchial reflex,
but they did not carried out upper GI endoscopy.This fact decline their results.
Response
We accept this as a limitation of the study. A statement to that effect has been included in the
discussion (see page 11 of revised manuscript).
Response to reviewer 2
Major comment 1
In a recent paper Rosztóczy et al. published detailed evaluation of the pulmonary and esophageal function
in patients with asthma, which includes esophageal acid perfusion test as well. Since the subject of their
paper has a significant overlap with the submitted manuscript, it seems to be essential to discuss those
results. In contrast to the results of this manuscript, they were unable to show significant changes in the
FEV1 for esophageal acid perfusion itself: However, during combined acid perfusion + metacholine
challenge asthmatic patients reached the significant FEV1 decrease at lower metacholine dose, than
during the standard metacholine test. (Rosztóczy A, Makk L, Izbéki F, Róka R, Somfay A, Wittmann T.
Asthma and gastroesophageal reflux: clinical evaluation of esophago-bronchial reflex and proximal
reflux. Digestion. 2008;77(3-4): 218-24. Epub 2008 Jul 19.)
Response
A paragraph addressing this has been included in the discussion (see page 10 of revised
manuscript).
“A similar study investigating 43 patients with bronchial asthma failed to show a significant
reduction in FEV1 during/ following esophageal acid perfusion test but that patients with
bronchial hyperreactivity, demonstrated by methacholine challenge showed a decrease. They
suggested that the esophago-bronchial reflex studied was not present in patients without bronchial
hyperreactivity.”
Major comment 2
Methods section – page 5 – In the referred article (24) DeMeester and Johnson discussed their results
obtained single channel intraesophageal pH monitoring. Why did authors applied this for their
measurements obtained by the proximal pH sensor. It is well known that those parameters are not
applicable for proximal reflux. What was the actual cut-off value for the determination of pathological
proximal reflux in this study?
Response
Using the article by Demeester and Johnson as reference for methodology, we performed 24 hour
pH monitoring in healthy subjects (published as Amarasiri DL, Pathmeswaran A, Dassanayake AS,
de Silva AP, Ranasinha CD, de Silva HJ. Esophageal motility, vagal function and gastroesophageal
reflux in a cohort of adult asthmatics. BMC Gastroenterology 2012, 12:140) and used their values
as cut-off criteria for classification of abnormal distal and proximal reflux.
This data has now been included in the paper as Table 1 (see table 1 below) and the reference cited
(see page 6 of revised manuscript).
Table 1 - Acid exposure in the oesophagus in healthy volunteers
Proximal sensor (n=16)
Distal sensor
th
Parameter
Median (range) 95 percentile Median (range)
Total time pH<4 (%)
Upright time pH<4 (%)
Supine time pH<4 (%)
Total no. of reflux episodes
No. of episodes  5 min
Duration of longest reflux,
min
DeMeester score
(n=26)
95th percentile
0.07 (0-0.4)
0.02 (0-0.4)
0.12 (0-0.5)
6.0 (0-25.0)
0 (0-1)
0.9 (0-7.8)
0.4
0.4
0.5
25.0
1.0
7.8
0.4 (0.08-1.5)
0.2 (0-3.0)
0.4 (0.07-2.8)
18.0 (2.7-94.2)
0 (0-2)
1.9 (0.5-20.6)
1.5
3.0
2.8
94.2
2.0
20.6
0.95 (0.2-4.8)
4.8
3.35 (0.7-11.5)
11.5
Major comment 3
Methods and results section – page 6 and 7 – The authors used concentrated lime juice in their
experiments for the acid perfusion test. In the manuscript they write: “concentrated lime juice (pH 2-3)
alternatively, at a rate of 2mL/min for 10 minutes” Why has that been chosen? Please add references!
What was the actual pH (pH 2 and means 10 times higher H+ ion concentration than pH 3)?
Why do they think that acidity itself and not something else is responsible for the observed changes (lime
juice has a number of components)?
Furthermore, saline and lime juice looks and smells differently. How were the subjects and the examiners
blinded to eliminate the obvious placebo effect? Written data in this section do not suggest doubleblinded testing!
Has they performed any cross over tests? Normally, acid perfusion tests require at least 2-2 infusions of
acid and saline in a double blinded manner.
Why did they apply 30 minutes resting period between esophageal infusions and vagal/pulmonary
function testing?
Response
The argument for using lime juice was that it is more physiological. Most acid perfusion studies use
0.1N HCl which is a very strong acid level and will anyhow be able to elicit a response. Gastric juice
is also not a pure solution of acid, hence use of pure HCl is also not representative of what actually
happens in the gut. Therefore we cannot declare that the changes observed are due to acid alone.
We accept this as a limitation of the study.
The infusion rate was based on the acid perfusion study of Wu et al. 2000 (Wu DN, Tanifuji Y,
Kobayashi H, Yamauchi K, Kato C, Suzuki K, Inoue H. Effects of esophageal acid perfusion on
airway hyperresponsiveness in patients with bronchial asthma. Chest 2000, 118(6):1553-1556. This
has been now cited at the relevant place in the methodology (see page 7 in revised manuscript).
The actual pH was 3 (this has been adjusted in the methodology) (see page 7 in revised manuscript).
The subjects were in a supine position with eyes closed. We admit that they may have been aware of
the solution perfused. We accept this as a limitation of the study.
Though normally, acid perfusion tests require at least 2-2 infusions of acid and saline in a double
blinded manner, in our study each subject was tested at baseline, then a single infusion of acid
followed by saline.
We followed the study of Lodi et al. (Lodi U, Harding SM, Coghlan HC, Guzzo MR, Walker LH.
Autonomic regulation in asthmatics with gastresophageal reflux. Chest 1997, 111(1):65-70) who
allowed 30 minutes of resting prior to measuring ANF tests at baseline. This has been now cited at
the relevant place in the methodology (see page 6 in revised manuscript).
In the present study, ANF testing was done immediately following acid perfusion. We have now
clarified this statement in the methodology (see page 6 in the revised manuscript) and in the
abstract (see page 2 in revised manuscript). We apologize for the oversight.
Major comment 4
The asthma severity of the GERD positives and negatives seems to be different. Please add additional
statistics to clarify this!
Response
As demonstrated in table 1 : Of the 20 asthmatics who scored less than cut-off for GERD
symptoms; 17 had mild asthma (intermittent or persistent) and 3 had moderate or severe persistent
asthma.
Of the 20 asthmatics who scored more than the cut-off for GERD symptoms 14 had mild asthma
and 6 had more severe asthma.
The numbers were too small to compare these differences statistically.
Major comment 5
Authors state, that the difference between the mean FEV1 values after saline and acid perfusion is
0.1L, while the standard error is 0.4L for this parameter. I can hardly believe, that the statistical
evaluation would show such a high (p<0.001) level of significance. Furthermore, authors stated
different SE values in the text 2.7±0.1L and 2.6±0.1L. So, which is the real SE? Please provide the
exact FEV1 values of the studied patients in a separated data file for repeated statistical evaluation.
This comment relates to the data presented in table 4. The data presented here are as mean (SD).
By an oversight has not been stated in the table. This has now been corrected.
The analysis of the data in table 4was done by means of Repeated Measures Analysis which gave
the P values shown in the table. Therefore it is not possible to get a significant P value using other
summary statistics.
We accept that there is a difference in the values quoted in the text and the table. We have removed
the description from the text (as this is a repetition of the data in the table). We apologize for the
oversight.
Major comment 6
Control subjects were not used. Please add the following groups of controls (healthy subjects and nonasthmatic GERD patients).
Response
We accept that this is a limitation of the study. This has been addressed in the discussion (see page
11 of revised manuscript) as follows.
“A recent study reporting that esophageal hypersensitivity could be induced and maintained by
repeated short duration acid infusion at physiological pH levels (pH 1.8-4) in healthy subjects
confirms this statement [37]. Therefore, the response of healthy controls to acid perfusion was not
assessed in the present study. The lack of data of healthy subjects and non-asthmatic GERD
patients is a limitation. “
Minor comment 1
The introduction should be shortened.
Response
Introduction has been shortened (see pages 4 and 5 of revised manuscript).
Minor comment 2
Page 5 –”Stolkholm” is correctly Stockholm.
Response
This has been corrected in the manuscript (see page 5 of revised manuscript).
Minor comment 3
Page 5 and 6 – please use “LES” instead of “LOS” since always “esophagus” is used in the manuscript.
Response
This has been corrected in the manuscript (see page 5 of revised manuscript).
Minor comment 4
Page 6 – authors state: “The day following removal of the pH catheter, following an overnight fast, a
feeding tube (4mm diameter) was used to deliver acid to the distal esophagus. The tube was inserted to 15
cm above the upper border of the lower esophageal sphincter (LES).” In fact, 15 cm above the LES
should not be considered the distal but rather the middle (or the lower part of the proximal) esophagus.
Later in the discussion they state the following: “In the present study, the esophageal catheter was placed
at 10 cm above the LES.” So what was the real catheter position? Please clarify!
Response
The real catheter position was 15cm. This has been corrected in the discussion (see page 12 of
revised manuscript). We accept that the position of 15cm above LES be considered as lower part of
the proximal esophagus.
Minor comment 5
Page 6 – authors state: “Oral asthma drugs were withheld for 24 hours and inhaled drugs for 8 hours prior
to the study, allowing inhaled or nebulized beta-2 agonists on an as-required basis.” Did they exclude
those patients (or at least delay their testing) who needed such medication in the last 8 hours?
Response
Only 2 subjects needed such medication and they were tested once they were stable.
Minor comment 6
In table 2. data shown as “mean (SE)” instead of “mean±SE”. Shince the latter form is used in other sites
of the manuscript please correct!
Response
This has been corrected in the manuscript (see table 2).