Bronchial hyper-responsiveness in preterm-born subjects: A systematic review and meta-analysis

Background : Preterm-born survivors have increased respiratory symptoms and decreased lung function, but the nature of bronchial hyper-responsiveness (BHR) is unclear. We conducted a systematic review and meta-analysis for BHR in preterm-born survivors including those with and without chronic lung disease in infancy (CLD) comparing results to term-born subjects. Methods : We searched eight databases up to December 2016. Included articles compared BHR in preterm-born and term-born subjects. Studies reporting BHR as decreases in forced expiratory volume in 1 second (FEV 1 ) after provocation stimuli were included. The analysis used Review Manager V5.3. Results : From 10 638 titles, 265 full articles were screened, and 28 included in a descriptive analysis. Eighteen


Pl e a s e n o t e:
C h a n g e s m a d e a s a r e s ul t of p u blis hi n g p r o c e s s e s s u c h a s c o py-e di ti n g, fo r m a t ti n g a n d p a g e n u m b e r s m a y n o t b e r efl e c t e d in t his ve r sio n.Fo r t h e d efi nitiv e ve r sio n of t hi s p u blic a tio n, pl e a s e r ef e r t o t h e p u blis h e d s o u r c e.You a r e a d vis e d t o c o n s ul t t h e p u blis h e r's v e r sio n if yo u wi s h t o cit e t hi s p a p er.
Thi s v e r sio n is b ei n g m a d e a v ail a bl e in a c c o r d a n c e wit h p u blis h e r p olici e s. S e e h t t p://o r c a .cf. a c. u k/ p olici e s. h t ml fo r u s a g e p olici e s.Co py ri g h t a n d m o r al ri g h t s fo r p u blic a tio n s m a d e a v ail a bl e in ORCA a r e r e t ai n e d by t h e c o py ri g h t h ol d e r s .
long-standing structural consequences of preterm birth, as evidenced by smooth muscle extension into the smaller airways well beyond that observed in term-born infants, especially where the infant has been diagnosed with chronic lung disease of prematurity (chronic lung disease [CLD], also called bronchopulmonary dysplasia, BPD). 5,6Alternatively, there are limited data suggesting that an active pro-inflammatory neutrophilic status 7 and oxidant process may be continuing. 8e nature of the airway narrowing observed in preterm-born children can be tested using pharmacological stimuli, for example directly acting on smooth muscle cells' receptors or indirectly via the release of mediators by pro-inflammatory airway cells. 9This process of assessment of bronchial hyper-responsiveness (BHR) has been extensively used in asthma, although it is not as discriminatory as expected as many asymptomatic subjects may have increased BHR and those with the disease may not. 10Nevertheless, BHR by both direct (eg histamine, methacholine) and indirect (eg exercise, mannitol) means has been used in preterm-born survivors to try to elucidate the potential mechanisms underlying the airway obstruction observed.The current data do not conclusively confirm that BHR is increased in preterm survivors; furthermore, it is even less clear if direct agents have greater or similar effects to indirect agents.Thus, we conducted a systematic review and meta-analysis to identify if: 1. BHR was increased after preterm birth when compared to term-born controls.2. BHR was increased in preterm-born subjects who had CLD compared to term-born subject 3. Any increase in BHR was due to responses to (a) direct or (b) indirect stimuli.

| METHODS
We used methodologies and data from our three previous systematic reviews. 3,11,12As data for BHR are often not reported in titles or abstracts, we combined several approaches.We (a) re-ran the initial searches for articles reporting FEV 1 in preterm-born subjects 3

| Eligibility criteria
Studies on BHR in preterm-born subjects of any age, (adults and children), with or without CLD, and of any gender were included.
Randomized and nonrandomized intervention studies, prospective and retrospective case-control studies, and prospective and retrospective cohort studies were included.Preterm defined as birth <37 weeks' gestation and term as birth ≥37 weeks' gestation.Studies which recruited on the basis of birthweight were included if they reported gestational age and all subjects in the study group were preterm and the control group were term-born, or the birthweight cut-off for the preterm group was <1501 g.Authors' definitions of CLD and what constituted BHR, and all methods of assessing BHR were accepted.However, only studies reporting a change in FEV 1 after a challenge were included.
Studies in all languages from all countries were considered.

| Study selection
Searches were conducted in December 2016 and January 2017.
Two reviewers (SJK and HC) independently screened each reference title and available abstracts, using the inclusion criteria.Complete manuscripts were obtained for those that met the inclusion criteria as judged by either reviewer.The two reviewers then screened the full manuscripts against the inclusion criteria.Where there was disagreement, a third reviewer (SK) made the final decision.

| Data collection process
SJK data extracted included articles.Authors of articles were contacted, where possible, for further details if the information was not in a format which enabled data extraction for inclusion in the systematic review.Multiple articles from the same cohort were reviewed by SK and SJK, and the article reporting the most complete data was included.

| Assessment of study quality and risk of bias
Study quality was assessed based on criteria from the Newcastle Ottawa criteria and the Cochrane risk of bias tool.The quality assessment sheet is shown in the Data S1.Each study was scored for representativeness of the cohort, appropriate selection of the nonexposed group, exposure ascertainment and demonstration that the outcome of interest was not present at the start of the study, outcome assessment and adequacy of follow-up by SJK.Minimum score was six and maximum score was 20.

| Outcome measures
Number of subjects with BHR (given by a defined change in FEV 1 after BHR challenge) in the premature group and in the premature CLD subgroup, compared with a term control group.

| Analysis of results
A formal meta-analysis was conducted for the studies which reported the number of subjects in the (a) preterm (with and without CLD) and control groups, and (b) preterm group with CLD and a term control group who had a positive BHR result as defined by the authors of each manuscript.Both meta-analyses were further analysed by dividing the subjects into the different methods of testing for BHR to separate out the effects of direct and indirect challenges.The results of all studies including those not in the meta-analyses are also presented descriptively.A sensitivity analysis was performed to investigate the effect of year of birth of the preterm-born subjects on BHR by dividing the preterm-born subjects into two groups (a) born on or after 1990 and (b) born before 1990.Where the subjects were born over a range of years, a mid-point was used.

| Statistical analysis
Statistical analyses were performed using Review Manager (RevMan) version 5.3. 13After initial exploration of the data, we used randomeffects meta-analyses to allow for heterogeneity.Heterogeneity was assessed using the I 2 statistic produced by RevMan.The following were used as a rough guide I 2 "0%-40% might not be important; 30%-60%: may represent moderate heterogeneity; 50%-90%: may represent substantial heterogeneity; 75%-100%: considerable heterogeneity". 14ere was large heterogeneity between the articles, due to a range of methods to assess BHR, variable outcomes measures and range of ages and gestations studied over a number of years.

| Studies selected and their characteristics
A total of 10 638 article titles were identified of which 265 full articles were screened for inclusion.Twenty-eight  met the inclusion FIGURE 1 Study selection results criteria (Figure 1). For te two studies that overlapped 24,26 the one reporting BHR as a proportion of responders 26 was included in the meta-analyses.Detailed demographics of included articles are shown in Table E1 (Data S1), and a summary of the demographics for the direct and indirect methods of assessing BHR is shown in Table 1a and b, respectively.Of the 28 articles reporting BHR: 1. 14 performed a methacholine challenge 2. 12 performed an exercise test 3. 1 performed a cold air challenge 4. 1 performed testing with hypertonic saline Many of the articles studied preterm-born subjects where prematurity was defined as being born at a gestation of <37 weeks, 15,18- 21,23,29,34,37,39,40,42 but there was heterogeneity in the groups as articles also sometimes only included extremely (<28 weeks' gestation) or very (≤32 weeks' gestation) 17,[24][25][26][27][28][31][32][33]35,36,41 or late (33 -36 weeks' gestation) preterm-born subjects. 38 Although th majority of studies studied randomly selected preterm subjects or studied cohorts, one study included preterm infants with congenital diaphragmatic hernia (CDH), but the authors provided data to us to only include the preterm and term control infants excluding CDH.22 Sensitivity analyses removing that study from all analyses marginally increased the odds ratios in favour of the preterm groups.In most of the included studies, BHR was tested as part of wider ranges of lung function assessments.

| Risk of bias across studies
The quality scores ranged from 9 to 18 (median 14).The studies not included in the meta-analyses had similar quality scores to those included in the meta-analyses.Included studies range from 11 to 16 (median 14) and not-included studies range from 9 to 18 (median 14.5).

| BHR in preterm group compared to term control group
Eighteen of the 28 included articles were included in a meta-analysis.
The results of the meta-analysis of the 18 articles are shown in Figure 2A; demographics are described in the Data S1 (Table E1).
The results for subjects who had a methacholine (direct) challenge or who had an exercise test (indirect challenge) are reported in Figure 2B and C, respectively, and demographics are shown in Table E1 (Data S1).The pooled estimates of OR (95% CI) for BHR in the preterm group was 1.89 (1.12, 3.19) P = 0.009 for the eight articles reporting results after a methacholine challenge and 2.59 (1.50, 4.50) P = 0.0007 for the eight articles reporting results after an exercise test.
OR were greater for studies where the preterm-born subjects were born on or after 1990 than studies where the preterm-born subjects were born before 1990.

| BHR in preterm group who had CLD in infancy compared to term control group
Fifteen of the 18 articles also compared BHR in preterm-born subjects who had CLD with term controls, 17,18 , 21, 24-27, 29, 31, 33-37,41 see Data S1 for demographics (Table E1).We performed a metaanalysis for nine of the 15 articles although the total number studied was small (Figure 3A).The definitions of CLD used by the authors' of the nine articles are reported in Table E2 (Data S1).The pooled estimates of OR (95% CI) for BHR in the preterm group who had CLD was 4.54 (2.68, 7.69), P < 0.00001.The results for the subjects who had a methacholine challenge or who had an exercise test are reported in Figure 3B and C, respectively.The pooled estimates of OR (95% CI) for BHR in the preterm group who had CLD was 4.35 (2.36, 8.03), P < 0.00001 for the four articles reporting results after a methacholine challenge and 5.13 (1.82, 14.47), P = 0.002 for the five articles reporting results after an exercise test.It should be noted that one study used different definitions for the preterm (PC 20 <4 mg/mL) and CLD (PC 20 <1 mg/ mL) groups to define a positive BHR response. 27

meta-analyses
6,39,40 Six articles reported BHR after a methacholine challenge, 17,19, 23-26 , 39 and four articles reported after an exercise test. 29,33,36,40The studies are described in detail in the Data S1 (Table E1).In general, all studies for both methacholine and exercise reported increases in BHR but the results were not always significantly different from included term-born controls.However, comparisons between articles were difficult due to heterogeneity.

| DISCUSSION
The results of our systematic review and meta-analyses suggest preterm-born subjects have greater BHR compared to term-born subjects, and differences are greatest for subjects who had CLD.
Furthermore, the commonest used direct agent, methacholine, and indirect method, exercise testing, both resulted in greater BHR in preterm-born survivors with and without CLD and preterm-born subjects with CLD compared to term controls.Of note was the variety of different stimuli and outcome measures used, making comparisons difficult between studies.
Respiratory symptoms in school-age preterm-born children are often inappropriately labelled as asthma; however, the underlying mechanisms are likely to be different to those in asthma.Previous studies 43 have suggested that studying BHR by both direct and  indirect means may help elucidate the underlying mechanisms and aid targeted therapy, for example anti-inflammatory or smooth muscle relaxants.However, despite the various methods used in the meta-analyses and descriptive analyses, the collated results strongly suggest that BHR is more prevalent in preterm-born group compared to term-born subjects, especially in those who had CLD-a finding similar to those with asthma.
As different mechanisms may potentially be identified using direct (assessing smooth muscle phenotypes and responses) and indirect methods (pathways of inflammatory mediators), to assess B, Number of subjects with BHR after a methacholine challenge in the premature group compared with term control group.C, Number of subjects with BHR after an exercise test in the premature group compared with term control group BHR, 9 we classified the studies using direct and indirect means to assess BHR.This is an important distinction to make as there is a suggestion that BHR responses may be different in asthma and in lung disease of preterm-born children: Kim et al 43 reported children with asthma responded to both methacholine and adenosine 5′-monophosphate but children with CLD only responded to the methacholine suggesting that continuing inflammation may not be a factor in preterm-born subjects.Methacholine is a direct method of assessing BHR by assessing bronchial smooth muscle response, and adenosine 5′-monophosphate is an indirect method of assessing BHR by pathways of inflammatory mediator release from airway mast cells.Therefore, as children with CLD only responded to the direct method, not indirect methods, it is possible inflammation may not be a factor in BHR of the CLD subjects in the study by Kim et al.This is clearly contradictory to two studies which reported increased neutrophilic inflammation in induced sputum and increased oxidant activity in exhaled breath condensate from children with CLD. 7,8 he results, however, are in agreement with reports of low exhaled nitric oxide in children with CLD. 44terestingly, separating the data on whether a direct or indirect method was used showed that both methods resulted in increased BHR in preterm-born subjects including the CLD group despite smaller numbers available for inclusion.Historically, it has been shown at autopsy that airway smooth muscle is both thicker and extends further down the airways in preterm subjects especially those dying from CLD when compared to matched term controls. 5,6wever, what happens in the current cohorts of survivors can only be speculative.Our data suggest that preterm-born subjects have increased responses to direct stimuli suggesting that treatment with as the smooth muscle phenotype may initially respond to bronchodilators but there is speculation that these cells may develop a more fixed unresponsive myofibroblast phenotype which may not respond to bronchodilator treatment. 45r data also showed that exercise also resulted in convincing differences between the preterm groups when compared to term controls.It is likely that preterm-born subjects have either structural abnormalities or airway inflammation or both, but the relevant studies to confirm or refute either are currently lacking.

| LIMITATIONS
As with all systematic reviews, we were limited by the data in included articles.The majority of articles examined BHR as part of a wider range of lung function tests, and there was disparity in the articles.Interpreting the results was complex as the articles reported a range of methods used to test BHR.Variable inclusion criteria and agents-both pharmacological and physiological-were used to assess and report BHR outcomes which included induction of wheeze, changes in spirometry and changes in oxygen saturation.Even when a change in FEV 1 was used, the definitions of a positive response varied.A number of articles were excluded as they not only defined BHR after testing with a provocative agent/ test, but they also included data on subjects with a low FEV 1 who had responded to an inhaled bronchodilator.In addition, articles reporting BHR in preterm-born subjects without a term-born control group and articles with preterm and term-born subjects combined in the results were excluded.Variable criteria were used to define CLD in articles included in the meta-analyses.We ideally would have liked to further analysis the CLD group by dividing the group into two: (a) CLD defined as a requirement for supplemental oxygen at ≥28 postnatal days and (b) CLD defined as a requirement for supplemental oxygen at ≥36 weeks postmenstrual age.However, the small number of subjects in the articles with the later definition meant this was not possible.Additionally, the subjects were born at a wide range of gestational ages over a number of decades when medical management has progressed-whether BHR is affected by the survival of the extremely preterm babies is unclear.

| CONCLUSIONS
This is the first comprehensive systematic review and meta-analysis which collated and reported the effect of preterm birth on later BHR.Suggesting an increased rate of BHR in preterm-born subjects compared to term-born subjects, differences were greatest for subjects who had CLD.Both direct (methacholine challenge) and indirect (exercise) challenges resulted in increased BHR suggesting that subgroups of preterm-born subjects could potentially benefit from anti-inflammatory and/or bronchodilator therapies.
; (b) adapted the initial search strategies to include additional keywords relating to BHR and ran them in eight databases; and (c) searched references in the included articles to identify additional papers reporting BHR in preterm-born subjects compared to term-born subjects (see Data S1 for protocol, search strategy, and data collection form).Eight databases were searched: EMBASE, Health Management Information Consortium (HMIC), MEDLINE, Medline in Process, Scopus, OpenSIGLE, CINAHL and Web of Science.Ethical approval was not required.

FIGURE 2 A
FIGURE 2 A, Number of subjects with bronchial hyper-responsiveness (BHR) in the premature group compared with term control group.

FIGURE 3 A
FIGURE 3 A, Number of subjects with bronchial hyper-responsiveness (BHR) in the premature group who had chronic lung disease (CLD) in infancy compared with term control group.B, Number of subjects with BHR after a methacholine challenge in the premature group who had CLD in infancy compared with term control group.C, Number of subjects with BHR after an exercise test in the premature group who had CLD in infancy compared with term control group

TABLE 1 (
a) Characteristics of the studies using direct bronchial provocation tests; (b) Characteristics of the studies using indirect bronchial provocation tests

TABLE 1 (
Continued) 7,8The evidence for the efficacy of inhaled corticosteroids in lung disease of preterm-born survivors is also limited, although, taken together with our data, there is a suggestion that subgroups of preterm-born subjects may benefit from inhaled corticosteroids, although they may not be effective if neutrophilic inflammation is confirmed.Role of targeting the leukotriene pathway is also poorly studied in preterm-born children with lung disease.Appropriate studies to evaluate the efficacy of both bronchodilators and inhaled corticosteroids are urgently required.
43e preterm group with and without CLD has increased BHR to both direct and indirect but we were not able to assess if both were present in the same individuals.Besides the study by Kim et al43(who assessed BHR by auscultation and oximetry), the study by Nikolajev and colleagues of moderately preterm-born children (mean gestation of 35 weeks) reported overlap between exercise, methacholine and cold air challenges as well as with responses to bronchodilators.