Investigating Fractional Exhaled Nitric Oxide in Chronic Obstructive Pulmonary Disease (COPD) and Asthma-COPD Overlap (ACO): A Scoping Review

Abstract

Chronic obstructive pulmonary disease (COPD) is the most common fixed airflow limitation. Individuals may present with the features of both asthma and COPD called asthma-COPD overlap (ACO) with more severity and worse health-related quality of life than COPD or asthma. One of the promising biomarkers that could be used in clinical practice to differentiate ACO from COPD is fractional exhaled nitric oxide (F_ENO). The role of Fractional exhaled nitric oxide (F_ENO) in COPD/ACO remains unknown. This scoping review aims to investigate the role of F_ENO measurement to differentiate COPD from ACO, to anticipate disease severity/progression and treatment response. A structured comprehensive literature search was performed in major databases including Medline, EMBASE, CINAHL, Cochrane Library, Web of Science, and BIOSIS from 2005 onwards. Thirty-eight studies were retrieved. Based on the synthesis of the reviewed literature, six themes emerged. Thirty-four articles covered more than one theme. From which, 24 articles were on modifying factors in F_ENO measurement, 18 on F_ENO in COPD compared with healthy subjects, and seven on F_ENO in ACO compared with COPD, 22 on F_ENO and disease severity/progression,12 on F_ENO and biomarkers, and eight on F_ENO and treatment response.

F_ENO measurement cannot be used alone in the clinical settings of COPD patients. Although F_ENO level is higher in ACO patients than COPD-only, it is still unclear if there is a F_ENO cut-off that can be used to make the diagnosis of ACO and/or to guide therapy with inhaled corticosteroids/glucocorticoids in COPD patients.

KEYWORDS:

• Fractional exhaled nitric oxide (F_ENO)
• chronic obstructive pulmonary disease (COPD)
• asthma-COPD overlap (ACO)

View correction statement:

Correction

Introduction

Chronic obstructive pulmonary disease (COPD) is the most common obstructive pulmonary disease (1, 2). The cause of COPD is chronic exposure to noxious particles or gases, mainly tobacco smoke and is not generally recognized below the age of 40 (2–5). COPD is an inflammatory disease with persistent, progressive, and incomplete reversible airflow limitation, defined by the post-bronchodilator ratio of forced expiratory volume in 1 second over forced vital capacity (FEV1/FVC) below 0.7, or below the lower limit of normal (6, 7).

Individuals may present with clinical features of both asthma and COPD. This condition is called asthma-COPD overlap syndrome (ACO), as reflected in the Global Initiative for Asthma/Global Initiative for Chronic Obstructive Lung Disease (GINA/GOLD) statement (8) and other guidelines and literature (6, 9–11). The prevalence of ACO has been reported as low as 15% and as high as 60% according to different population samples, age groups, and definitions (6, 12). There is still no consensus on a definition of ACO. However, compared with asthma or COPD alone, there is an association between ACO using different definitions and more frequent exacerbations (13, 14), worse health-related quality of life (13, 15), increased hospital admissions (14, 16), and higher health care costs (17). To date, the diagnosis of ACO is based on questionnaires and doctor’s personal opinion as well as defining some minor and major criteria, but there is no agreement regarding these criteria (1, 6, 9). Finally, we know very little about the treatment of ACO since these patients have generally been excluded from both asthma and COPD studies (6, 18, 19) and drug trials (1, 20).

There is a need for a biomarker that could be used in clinical practice to differentiate ACO from COPD. The measurement of fractional exhaled nitric oxide (F_ENO) can assess airway inflammation, and thus, the American Thoracic Society clinical practice guideline recommends its use to manage and monitor asthma (21). However, the exact role of F_ENO in patients with ACO and COPD remains to be defined (4). F_ENO is produced in the catalysis of nitric oxide synthase in different kinds of respiratory epithelial cells, inflammatory cells, and vascular endothelial cells. It is used as a known marker of airway hyperresponsiveness, the total number of inflammatory cells in the airways, eosinophilic airway inflammation, and T-helper cell 2 (Th2)-mediated airway inflammation (22–24). It is measured via a fast, noninvasive, reproducible, and easy way in close to real time by utilizing electrochemical detection, chemiluminescence, or laser spectroscopy devices (22, 25). A high level of F_ENO is associated with eosinophilic inflammation (21, 26). Therefore, F_ENO has been used clinically for detecting eosinophilic airway inflammation, monitoring airway inflammation in asthma, evaluating corticosteroid responsiveness, and as a management tool of asthma (21, 27). However, few studies have reported on the use of the F_ENO level for monitoring ACO patients undergoing inhaled corticosteroid (ICS) treatment (3). While there have been a number of preliminary studies on measuring F_ENO in COPD, literature defining the role of F_ENO and the practical cut-off value in patients with COPD and ACO are minimal (28).

As this topic of F_ENO in COPD covers a wide range of research questions and because of its exploratory nature, we decided to conduct a scoping review. The goal of this scoping review was to present an overview of the existing literature in a field of interest, i.e., F_ENO in COPD, and as well synthesize and aggregate findings from different studies. The research questions and the related themes are presented in Table 1. The composition of research on this topic will help researchers to find out the current state of the evidence and determine areas for future research.

Table 1. Themes, research questions, and the number of studies those were able to answer the related questions.

No.	Theme title	Research questions	Number of studies (%)
1	Factors modifying FENO Measurement	Which factors have been demonstrated to modify the level of FENO in COPD patients?	24 (63.15)
2	FENO in COPD or in COPD compared with Healthy Subjects	Does the FENO level increase in patients with COPD and is there an optimal cut-off value that may be useful in COPD versus healthy subjects?	18 (47.36)
3	FENO and Asthma-COPD Overlap (ACO)	Are there differences in FENO levels between patients with asthma COPD overlap (ACO), i.e., with asthma features and patients with COPD-only?	7 (18.42)
4	FENO and disease severity and/or progression	Are FENO values associated with disease severity and/or disease progression in patients with COPD?	22 (57.89)
5	FENO and Inflammatory Biomarkers	Are FENO values associated with changes in the level of inflammatory biomarkers such as blood/sputum eosinophils and/or IgE in patients with COPD?	12 (31.57)
6	FENO and Treatment Response	Whether increased or decreased FENO level has an influence on treatment response, especially ICS/GCS therapy with or without inhaled bronchodilators in patients with COPD.	8 (21.05)

F_ENO: fractional exhaled nitric oxide; COPD: chronic obstructive pulmonary disease; ACO: asthma-COPD overlap; IgE: immunoglobulin E; ICS: inhaled corticosteroid; GCS: glucocorticoid.

Methods

This study was conducted as per the methodology outlined in the Joanna Briggs Institute Reviewers’ Manual (29) and reported as per the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA-P) statement (30). It is also according to the Levac et al.’s (31) and Arksey et al.’s (32) framework for scoping review.

According to Arksey’s (32) recommendations, the scoping review process included the following six key steps: (1) detection of the research question, (2) detection of relevant studies, (3) study selection, (4) charting the data, (5) gathering, summarizing, and reporting the results, and (6) consultation which is optional. The full methodology (protocol) of this scoping review has been published in BMJ Open journal (33).

Eligibility criteria, participants, and type of study

Studies of all languages were considered, from 2005 onwards and including n > 10, those using fractional exhaled nitric oxide measurement and having patients diagnosed as COPD and/or asthma-COPD overlap (ACO). The randomized clinical trials (RCT), cohorts, cross-sectional, and longitudinal studies were only included.

Terminology of fractional exhaled nitric oxide in presenting results

According to the recent recommendation (34), this format, F_ENO, was used for the fractional concentration of exhaled nitric oxide in the gas phase (ppb) in presenting the results. The exhalation flow rate (if indicated in included articles in this scoping review) was given as a subscript in mL·s⁻¹. For instance, a flow rate of 50 mL·s⁻¹ was written F_ENO50 (34). If there was no direct or indirect evidence regarding the flow rate (mL s⁻¹) in the included article, we only used F_ENO. This format, F_nNO, was used for the fractional concentration of nasally aspirated/exhaled nitric oxide in presenting the results. The flow rate of the aspirated nitric oxide (if indicated in the included articles in this scoping review), usually 5 mL·s⁻¹, was given as a subscript in mL·s⁻¹ (e.g., F_nNO5) (34). If there was no direct or indirect evidence regarding nasally exhaled/aspirated flow rate (mL·s⁻¹) in the included article, we only used F_nNO.

Information source (databases), literature search, and search strategy

A structured comprehensive literature search was conducted in major databases including Medline (via OvidSP), EMBASE (via OvidSP), CINAHL (via EBSCO host), Cochrane Library (via Wiley Online), Web of Science, BIOSIS Previews, and BIOSIS Previews Archives on 26 June 2016. The inception of database searches was without date limitation but, then it was limited to the year 2005 onwards as this was the year when the ATS/ERS (American Thoracic Society/European Respiratory Society) guideline concerning fractional exhaled nitric oxide and its measurement was published (35). We performed the new search on 27 June 2017, as updated literature search. No systematic and scoping reviews were retrieved through these database searches (previous and current search). We used a variety of keywords/text words and database subject heading such as [COPD OR Chronic Obstructive Lung Disease OR Chronic Obstructive Pulmonary Disease OR Emphysema OR Chronic Bronchitis OR ACOS OR Asthma-COPD Overlap Syndrome OR Concomitant asthma] AND [FeNO OR Fractional Exhaled Nitric Oxide]. The search was done regardless of positive or negative results of studies. Details of the search strategy are available as Supplementary files.

Study selection

Two reviewers (SMY Mostafavi-Pour-Manshadi and Nafiseh Naderi) independently screened the title and abstract of retrieved articles from the database searches. Then, the full text of potential articles retrieved from first screening was investigated as the second screening. Discrepancies were solved by reaching the consensus between two reviewers according to the criteria eligibility. If the two reviewers could not reach the consensus concerning the specific article(s) or both were suspicious about including/excluding the articles, these papers were reviewed by the third reviewer (Jean Bourbeau) and the issue was solved.

Data extraction

Data collection/extraction was done by using a designated data extraction form and gathered electronically. We used PICOS (30, 36) approach for designing the form and extracting data, which was developed from our research questions. The form was reviewed and revised by the reviewers after completing to reach the consensus among reviewers. Data extraction was independently done by two reviewers (SMY Mostafavi-Pour-Manshadi and Nafiseh Naderi). The data included study title, first author’s name, publication year, the country of conducted study, the name of the journal, the purpose of the study, fractional exhaled nitric oxide measuring method, sample size, sample description, setting description, and outcomes. The information from the studies was summarized by producing descriptive summary tables.

Quality assessment of included studies

We did not appraise methodological quality or risk of bias of the included articles due to nature of this work, which is scoping review (29). This approach is consistent with scoping reviews of clinical topics (37, 38).

Results

According to the PRISMA diagram (Figure 1), 2,596 articles from 2005 onward were screened, 2,516 articles were excluded due to irrelevant either topic or abstract or according to the exclusion criteria. Finally, 80 articles were selected for review of the full text and 38 articles were chosen to be included in the final scoping review. Based on the synthesis of the reviewed literature, six themes on fractional exhaled nitric oxide and COPD emerged. Thirty-four articles covered more than one theme. Details of all included studies are available in the Supplementary file (Table S1).

Figure 1. PRISMA flow diagram.

Factors modifying fractional exhaled nitric oxide measurement

The studies (N = 24, 63.15%) included in this theme are presented in Table 2. One of the most frequently tested factors was cigarette smoking. Most studies showed a decreased F_ENO50 with current smoking (39–46); only one study (47) reported an increased (F_ENO) and one (48) showed no association. Relationship with inhaled corticosteroid (ICS)/systemic corticosteroid (GCS) (either intravenous or oral therapy) was also frequently tested. Five studies (3, 39, 49–51) showed a decreased F_ENO50 and five studies (43, 46, 48, 52, 53) found no association (F_ENO50 (43, 46, 52, 53), F_ENO (48)). Two studies (54, 55) reported decreased F_ENO50 with an inhaled combination of ICS/long-acting beta2 agonist (LABA) and two studies (41, 56) showed no association (F_ENO (41), F_ENO50 (56)). Only one study (56) reported on exercise (6-minute walk test) showing an association with a decreased F_ENO50. One study reported and showed that sodium bicarbonate mouth rinse could decrease F_ENO50 (57). All the studies that assessed exacerbations (3/3) reported that exacerbation is associated with an increased F_ENO50 (51, 58)/F_ENO (41). Cold weather and viral infection have been reported as potential factors to increase F_ENO (41). Finally, two studies reported on sex and F_ENO50, one showing a decreased F_ENO50 in females (44) and one no association (39); and three studies reported on age and F_ENO50, one showing an increased F_ENO50 with older age (51), one no association (43), and one negative association (59).

Table 2. Factors modifying fractional exhaled nitric oxide measurement^a.

References	Decreased	Increased	No effect/association	Comments
Bhowmik et al. (41)	Smoking	Cold weather/viral infection, exacerbation	ICS/LABA	Current COPD smokers versus ex-smokers Increased FENO in cold weather (October to December) could be due to viral infections.
Liu et al. (48)	–	–	Smoking, GCS	Smoking was reported in both COPD and healthy subjects; FENO
de Laurentiis et al. (42)	Smoking	–	–	Current COPD smokers versus ex–smokers, FENO50
Kunisaki et al. (50)	ICS	–	–	ICS therapy to COPD patients was prescribed for 4 weeks, FENO50
Roy et al. (44)	Smoking, Sex (female)	–	–	COPD smoker versus COPD ex-smoker, COPD women versus COPD men, FENO50
Dummer et al. (49)	GCS	–	–	Oral prednisone was prescribed for 3 weeks. Current smokers were excluded; FENO50
Antus et al. (39)	Smoking, GCS	–	Sex (male or female)	Exacerbated COPD smokers versus ex-smokers Systemic (IV or oral) corticosteroids were prescribed in all exacerbated patients in the hospital. Tend to have decreased FENO50 with GCS therapy at discharge.
Lehouck et al. (43)	Smoking	–	ICS, Age	Current smokers in both COPD patients and healthy controls. Prescribed ICS (39% of COPD patients) among different GOLD stages (I-IV); FENO50
Tilemann et al. (45)	Smoking	–	–	Current COPD smokers versus non-smokers (never smokers and ex-smokers), FENO50
Rouhos et al. (57)	Sodium bicarbonate	–	–	Both COPD and healthy subjects whom were prescribed sodium bicarbonate to rinse their mouth; FENO50
Bazeghi et al. (40)	Smoking	–	–	Current COPD smokers versus ex-smokers, FENO50
Akamatsu et al. (54)	ICS/LABA	–	–	ICS was prescribed twice daily for 12 weeks; FENO50
Antus et al. (52)	–	–	GCS, LABA, LAMA	Exacerbated COPD patients were prescribed systemic (IV or oral) corticosteroids and bronchodilators during hospital stay. No difference between low (<27 ppb) and high FENO50 (≥ 27 ppb) regarding GCS and bronchodilator therapy.
Xia et al. (58)	–	Smoking, exacerbation	–	Current COPD smoker versus non-smoker Exacerbation effect was considered in mixed group of subjects including healthy, stable COPD, asthma and non-asthma; FENO50
Rawy et al. (59)	Age	–	–	Negative association between FENO50 and age
Ishiura et al. (74)	–	–	ICS/LABA	ACO patients were prescribed ICS/LABA for 12 weeks; FENO50
Santini et al. (46)	Smoking	–	ICS	COPD current smokers versus COPD ex-smokers No difference in FENO50 values between COPD patients on ICS therapy and those not on ICS therapy.
Logotheti et al. (51)	ICS	Older age, exacerbation	–	COPD patients were prescribed an inhaled corticosteroid (ICS) for maintenance treatment. Older COPD patients (69 ± 6 years) versus younger patients (66 ± 7 years). A significant increase in exacerbations during previous year in COPD patients with elevated FENO50 levels.
Amer et al. (72)	–	Bronchodilator	–	FENO50 was measured before and 15 minutes after short inhaled bronchodilator therapy.
Ji et al. (55)	ICS/LABA	–	–	ACO patients were prescribed ICS/LABA for maintenance treatment; FENO50
Huang et al. (56)	6MWT (exercise)	–	–	FENO50 was measured after exercise test in COPD patients.
Kobayashi et al. (53)	–	–	ICS	ACO patients with prescribed ICS versus non-ACO with prescribed ICS, FENO50
Feng et al. (3)	ICS	–	–	COPD patients were prescribed ICS, 3 times per day for 6 months. FENO50 was measured before and after ICS therapy.
Zhao et al. (71)	–	–	Bronchodilator	FENO50 was measured before and 15 minutes after inhaled short-acting bronchodilator.

F_ENO: fractional exhaled nitric oxide; ICS: inhaled corticosteroid; LABA: long-acting beta agonist; GCS: glucocorticosteroid; COPD: chronic obstructive pulmonary disease; IV: intravenous; ppb: parts per billion; GOLD: Global Initiative for Chronic Obstructive Lung Disease; ACO: asthma-COPD overlap; 6MWT: 6-minute walk test.

^a– = Not mentioned, for expressing the F_ENO, the exhalation flow rate (if indicated) was given as a subscript in mL·s⁻¹. For instance, a flow rate of 50 mL·s⁻¹ was written F_ENO50. If there was no direct or indirect evidence regarding the flow rate (mL·s⁻¹), we only used F_ENO.

Fractional exhaled nitric oxide in COPD or in COPD compared with healthy subjects

The studies (N = 18, 47.36%) included in this theme are presented in Table 3 (COPD compared with healthy subjects). Thirteen studies reported an increased F_ENO50 (3, 25, 28, 42, 46, 51, 55–58, 60)/F_ENO45 (61)/F_ENO (48) in COPD and/or ACO, 2/2 studies in ACO (3, 55), and 11/16 in COPD (25, 28, 42, 46, 48, 51, 56–58, 60, 61). In contrast, only one study (45) showed a reduction in F_ENO50 levels in COPD patients compared with those with no airway disease and four studies (40, 43, 62, 63) reported no change/difference in F_ENO50 (40, 43, 62)/F_nNO5 (63) levels among COPD patients or between COPD patients and age-matched healthy subjects. For most studies, COPD and/or ACO patients were compared with healthy subjects. In one study (51), F_ENO50 was compared within COPD GOLD groups showing an increase from GOLD A to GOLD D. There were no cut-off values assessed to differentiate COPD and healthy subjects.

Table 3. Fractional exhaled nitric oxide in COPD or in COPD compared with healthy and in ACO compared with COPD subjects^a.

COPD compared with healthy
References	InCOPD or compared with healthy^b	Optimal cut-off value^c	Comments
Foschino Barbaro et al. (61)	Increase	–	COPD patients compared with healthy control subjects and reversible COPD compared with non-reversible COPD, FENO45
Liu et al. (48)	Increase	–	COPD patients compared with healthy subjects, FENO
de Laurentiis et al. (42)	Increase	–	COPD patients compared with healthy subjects: Not significant baseline mean level of FENO50 but significant FENO50 mean coefficient of variability (CoV).
Beg et al. (60)	Increase	–	COPD patients compared with healthy subjects, FENO50
Lehouck et al. (43)	–	–	COPD patients compared with age-matched healthy control subjects, FENO50
Tilemann et al. (45)	Decrease	–	COPD patients compared with subjects with no airway obstruction, FENO50
Rouhos et al. (57)	Increase	–	COPD patients compared with healthy subjects, FENO50
Bazeghi et al. (40)	–	–	COPD patients only (severe emphysema, chronic bronchitis, frequent exacerbations, low body mass), FENO50
Donohue et al. (28)	Increase	–	COPD patients with different phenotypes, especially those with concomitant asthmas, FENO50
Xia et al. (58)	Increase	–	COPD patients compared with healthy subjects and also exacerbated COPD compared with stable COPD. No association in FENO50 levels between stable COPD and healthy subjects
Durmaz et al. (63)^f	–	–	COPD patients at admission (exacerbation) or before discharge, FnNO5
Santini et al. (46)	Increase	–	COPD ex-smoker compared with healthy ex-smoker, FENO50
Arif et al. (62)	–	–	COPD patients, FENO50
Alcazar-Navarrete et al. (25)	Increase	–	COPD patients compared with non-smoker healthy controls, FENO50
Logotheti et al. (51)	Increase	–	COPD patients with GOLD stages (ABCD), FENO50
Ji et al. (55)	Increase	–	ACO patients compared with healthy subjects, both before and after treatment (ICS/LABA), FENO50
Huang et al. (56)	Increase	–	COPD patients compared with healthy subjects, FENO50
Feng et al. (3)	Increase	–	ACO patients compared with healthy subjects, FENO50
ACO compared with COPD
References	In ACO compared with COPD^d	Optimal cut-off value^e	Comments
Tamada et al. (27)	Increase	–	Higher FENO50 values in COPD with asthma-like airway inflammation (ACO) among COPD patients
Alcazar-Navarrete et al. (25)	Increase	19 ppb	Higher FENO50 values in ACO patients than those with other COPD phenotypes
Chen et al. (4)	Increase	22.5 ppb	Higher levels of FENO50 in ACO group than COPD group
Goto et al. (65)	Increase	–	Higher levels of FENO50 in ACO than those with COPD alone
Kobayashi et al. (53)	Increase	23 ppb	Higher FENO50 levels in ACO than non-ACO (among COPD patients).
Deng et al. (64)	Increase	29 ppb	Higher levels of FENO50 in ACO than COPD patients
Cosío et al. (67)	Increase	–	Higher levels of FENO50 in ACO than COPD patients

F_ENO: fractional exhaled nitric oxide; COPD: chronic obstructive pulmonary disease; ACO: asthma-COPD overlap; F_nNO: fractional concentration of nasally aspirated/exhaled nitric oxide; CoV: coefficient of variability; FEV1: forced expiratory volume in one second; ICS: inhaled corticosteroid; LABA: long-acting beta agonist; ppb: parts per billion; GOLD: Global Initiative for Chronic Obstructive Lung Disease

^aFor expressing the F_ENO, the exhalation flow rate (if indicated) was given as a subscript in mL·s⁻¹. For instance, a flow rate of 50 mL·s⁻¹ was written F_ENO50. If there was no direct or indirect evidence regarding the flow rate (mL·s⁻¹), we only used F_ENO.

^b In COPD compared with healthy: – = no difference/change.

^cOptimal cutoff value: – = not mentioned.

^dACO definition is different among the studies using major and minor criteria (25, 65) or either Global Initiative for Asthma (GINA)/Global Initiative for Chronic Obstructive Lung Diseases (GOLD) joint document 2014 (64) or 2015 (updated document) (53) or positive bronchodilator test and history of asthma (4) or history of asthma and smoking (≥ 20 pack-years) and chronic airflow limitation or diagnosed COPD and elevated blood eosinophil count (67) or high levels of fractional exhaled nitric oxide (F_ENO50 > 35 ppb) or immunoglobulin E (IgE ≥173 IU/mL) in diagnosed COPD (27). Details are available in the Supplementary file (Table S1).

^eOptimal cutoff value: – = not mentioned.

^fThe flow rate of the nasally exhaled/aspirated nitric oxide (F_nNO) was given as a subscript in mL·s⁻¹ (F_nNO5).

Fractional exhaled nitric oxide and asthma-COPD overlap (ACO)

The studies (N = 7, 18.42%) included in this theme are presented in Table 3 (ACO compared with COPD). Higher levels of fractional exhaled nitric oxide in ACO patients compared with those with COPD-only were observed in all studies. Four studies (4, 25, 53, 64) showed that F_ENO50 can be useful to differentiate ACO from COPD and introduced optimal cut-off values of 19, 22.5, 23, and 29 ppb with a sensitivity of 68%, 70%, 73%, 80%, and a specificity of 75%, 75%, 68.2%, and 73%, respectively. The area under the curve (AUC) for the cut-off 19, 22.5, 23 and 29 ppb was 0.79, 0.78, 0.74, and 0.85, respectively (4, 25, 53, 64). On the other hand, one study (65) indicated that F_ENO50 measurement cannot differentiate ACO from COPD. In that study, although F_ENO50 level was higher in ACO, especially in former smokers, compared with COPD alone, the AUC for the cut-off was 0.63. One study (27) used the cut-off value of 35 ppb (proposed by other studies) to determine the prevalence of ACO among COPD patients. Using this cut-off value, the prevalence of 16.3% was reported for ACO among COPD patients.

The studies used not only different cut-off values but also different F_ENO50measurement devices, sample size and definitions of ACO subjects (Table S1). Concerning ACO definitions, history of asthma was included in all of these definitions, except one (27). Two studies used GINA definition (53, 64), from which, one (64) used the GINA-GOLD 2014 (8) and the other (53) used GINA-GOLD Diagnosis of diseases of chronic airflow limitation: Asthma, COPD and asthma-COPD overlap syndrome (ACOS). 2015. (66). Two studies used both major and minor criteria (25, 65). ACO was diagnosed if there was either one major criterion or two minor criteria. Major criteria were defined in one of these studies (25) as a previous history of asthma/wheezing outside chest infections or an increased FEV1 > 14% and >400 mL post-bronchodilator; in the other study (65) as a previous history of asthma and bronchodilator response (increased FEV1 ≥ 15% and 400 mL). Minor criteria (25, 65) were as follows: positive bronchodilator response defined as ≥12% and/or 200 mL gain in FEV1, elevated blood eosinophil count, elevated IgE levels, history of atopy, hay fever or history of sensibilization to neumoallergens. Other studies (4, 67) used one minor criterion (positive bronchodilator test) and history/diagnosis of asthma or one minor criterion (elevated blood eosinophils) only (in diagnosed COPD) or chronic airflow limitation and a smoking history ≥20 pack-years (in diagnosed asthma patients) to define ACO. One study (27) used high levels of fractional exhaled nitric oxide (F_ENO50 > 35 ppb) or elevated immunoglobulin E (IgE ≥173 IU/mL) as candidate markers of ACO in COPD.

Fractional exhaled nitric oxide and disease severity and/or progression

The studies (N = 22, 57.89%) included in this theme are presented in Table 4. Twelve studies were on disease severity, from which, six studies assessed the disease severity by GOLD (4, 25, 27, 28, 43, 62), five studies by exacerbations (41, 42, 52, 58, 63), and one by both (51). For those studies using either GOLD airflow obstruction severity (I-IV) or GOLD 2011 risk assessment (ABCD), only one study (51) reported an association between GOLD 2011 (ABCD) (increased F_ENO50 from GOLD A to D) and F_ENO50 levels while six studies (4, 25, 27, 28, 43, 62) showed no association. Moreover, in the only study (51) showing an association with GOLD 2011 (ABCD), it was shown that COPD patients with high F_ENO50 levels who were not on corticosteroid had a significant increase in hospital care (need for corticosteroids or bronchodilators or admission to the Intensive Care Unit) compared with those who had low F_ENO50 levels. For the studies using exacerbations to evaluate the severity of the disease (six studies), four studies (42, 51, 52, 58) showed that COPD patients with higher F_ENO50 levels had an increased frequency or number of exacerbations per year. On the other hand, two studies (41, 63) showed no association between F_ENO (41)/F_nNO5 (63) levels and exacerbations. Definition of exacerbation was different among studies. One study used event-based definition (42), another one (52) used symptom based or significant change in prescribed medication. Two studies (41, 63) used symptom based definition only and two studies (51, 58) did not define the exacerbation in their study. Details regarding these definitions are available in Table S1.

Table 4. Fractional exhaled nitric oxide and disease severity and/or progression^a.

References	Association			Comments^b
	Disease severity		Disease progression
	GOLD	Exacerbations	Pulmonary Function Tests
Bhowmik et al. (41)	–	N	N	No association between FENO and FEV1, FVC or exacerbation frequency
de Laurentiis et al. (42)	–	Y	N	No association between FENO50 and FEV1; exacerbation rate was significantly associated with FENO50
Kunisaki et al. (50)	–	–	N	No association between FENO50 and FEV1 or FVC
Beg et al. (60)	–	–	Y	Negative association between FENO50 and FEV1/FVC ratio
Dummer et al. (49)	–	–	N	No association between FENO50 and FEV1
Antus et al. (39)	–	–	Y/N	Positive association between FENO50 levels at admission (acute exacerbation) and the post-treatment increase in FEV1 and FEV1% predicted No association with FVC
Lehouck et al. (43)	N	–	N	No association between FENO50 and GOLD stages (I–IV) No association with FEV1
Akamatsu et al. (54)	–	–	N	No association between FENO50 and changes in FEV1 as well as other pulmonary physiological parameters
Antus et al. (52)	–	Y	–	Association between low FENO50 level and more exacerbations (increased number of exacerbation per patient-year) during the follow-up
Donohue et al. (28)	N	–	–	No association between FENO50 levels and GOLD stages (I–IV)
Xia et al. (58)	–	Y	N	Association between elevated FENO50 and acute exacerbations (compared with stable COPD) No association with FEV1 and FEV1/FVC
Tamada et al. (27)	N	–	N	No association between high and low FENO50 levels with GOLD stages (I–IV) No association between low (≤ 35 ppb) and high FENO50 (> 35 ppb) levels with pulmonary function tests (FVC, FEV1, and FEV1/FVC)
Rawy et al. (59)	–	–	N	No association between FENO50 and FEV1/FVC
Durmaz et al. (63)^c	–	N	–	No association between FnNO5 level at presentation or before discharge and exacerbated patients
Arif et al. (62)	N	–	–	No association between FENO50 and GOLD stages (I–IV)
Alcazar-Navarrete et al. (25)	N	–	–	No association between FENO50 and GOLD 2011 (ABCD)
Logotheti et al. (51)	Y	Y	–	Association between FENO50 and GOLD 2011 (ABCD). Association with the increase in a number of exacerbations
Chen et al. (4)	N	–	–	No association between FENO50 and GOLD stages (I–IV)
Ji et al. (55)	–	–	N	No association between FENO50 and FEV1% predicted in ACO
Huang et al. (56)	–	–	N	No association between FENO50 and FEV1% predicted or the FEV1 change secondary to 6MWT
Feng et al. (3)	–	–	Y	Negative association between FENO50 levels with FEV1% predicted and FEV1/FVC in ACO patients
Deng et al. (64)	–	–	N	No association between FENO50 with FEV1% predicted and FEV1/FVC in COPD and ACO

GOLD: Global Initiative for Chronic Obstructive Lung Disease; F_ENO: fractional exhaled nitric oxide; FEV1: forced expiratory volume in first second; FVC: forced vital capacity; COPD: chronic obstructive pulmonary disease; F_nNO: fractional concentration of nasally aspirated/exhaled nitric oxide; 6MWT: 6-minute walk test; ACO: asthma-COPD overlap.

^aDisease severity and progression: Y = Yes, N = No, – = Not mentioned, for expressing the F_ENO, the exhalation flow rate (if indicated) was given as a subscript in mL·s⁻¹. For instance, a flow rate of 50 mL·s⁻¹ was written F_ENO50. If there was no direct or indirect evidence regarding the flow rate (mL·s⁻¹), we only used F_ENO.

^bSubjects are COPD patients unless stated otherwise

^cThe flow rate of the nasally exhaled/aspirated nitric oxide (F_nNO) was given as a subscript in mL·s⁻¹ (F_nNO5).

Fifteen studies assessed disease progression. These studies used pulmonary function tests to assess disease progression. Twelve studies reported no association between F_ENO50 (27, 42, 43, 49, 50, 54–56, 58, 59, 64)/F_ENO (41) levels and pulmonary function tests while two studies showed an association (F_ENO50) (3, 60). One study (39) reported both the association and no association, i.e., an association between F_ENO50 with FEV1 and FEV1% predicted while no association between F_ENO50 and FVC (Table 4). From studies showing an association-only, one (3) study showed a negative association between F_ENO50 with both FEV1% predicted and FEV1/FVC while another study (60) showed a negative association between F_ENO50 levels and FEV1/FVC in COPD/ACO patients.

Fractional exhaled nitric oxide and inflammatory biomarkers (sputum/blood eosinophils and IgE)

The studies (N = 12, 31.57%) included in this theme are presented in Table 5. Ten studies reported on sputum eosinophils and fractional exhaled nitric oxide levels, 9/10 showing an increased F_ENO50 (3, 44, 49, 55, 59, 68–70)/F_ENO45 (61) and one study showing no association (F_ENO50) (71) between COPD/ACO patients with elevated sputum eosinophils and those without elevated sputum eosinophils. Elevated sputum eosinophils were defined in the studies as ≥2.5% (69) or ≥3% (70) or >3% (68). Other studies did not define sputum eosinophilia (3, 44, 49, 55, 59, 61, 70). Four studies reported on blood eosinophils and F_ENO50, Two studies (45, 59) showed a relationship between F_ENO50 levels with blood eosinophils (elevated blood eosinophil count not defined), and two studies (67, 69) reported no association (either elevated blood eosinophils ratio or count defined as blood eosinophil count ≥1% (69) or ≥200 eosinophils·μL⁻¹) (67). Four studies reported on IgE, all studies (3, 45, 55, 68) showed a relationship between IgE and F_ENO50levels.

Table 5. Fractional exhaled nitric oxide and inflammatory biomarkers^a.

References	Association			Optimal cut-off value	Comments^d
	Sputum eosinophils^b	Blood eosinophil^c	Serum IgE
Foschino Barbaro et al. (61)	Y	–	–	–	COPD patients with airway reversibility, FENO45
Roy et al. (44)	Y	–	–	–	Regardless of percentage differential or cell count, FENO50
Dummer et al. (49)	Y	–	–	–	Off-steroid FENO50 and sputum eosinophil percentage
Tilemann et al. (45)	–	Y	Y	–	Mixed group of patients (COPD, asthma, and patient with partial reversibility), FENO50
Soter et al. (70)	Y	–	–	19 ppb	Regardless of percentage/number. Both at exacerbation and discharge, FENO50
Rawy et al. (59)	Y	Y	–	–	Positive association, blood and sputum eosinophil percentage, FENO50
Chou et al. (68)	Y	–	Y	23.5 ppb	Patients with sputum eosinophilia compared with those without sputum eosinophilia, FENO50
Ji et al. (55)	Y	–	Y	–	Pre- and post-treatment (ICS/LABA) FENO50 levels with sputum eosinophils and serum total IgE.
Feng et al. (3)	Y	–	Y	–	ACO patients, FENO50
Cosío et al. (67)	–	N	–	–	Patients with elevated blood eosinophil compared with those without elevated blood eosinophil, FENO50
Gao et al. (69)	Y	N	–	17.5 ppb	Patients with elevated eosinophils either in sputum or in blood compared with those without elevated sputum/blood eosinophils, FENO50
Zhao et al. (71)	N	–	–	–	Before and after bronchodilator inhalation, FENO50

F_ENO: fractional exhaled nitric oxide; IgE: immunoglobulin E; COPD: chronic obstructive lung disease; ppb: parts per billion; ICS: inhaled corticosteroid; LABA: long-acting beta agonist; ACO: asthma-COPD overlap.

^aY = Yes, N = NO, – = Not mentioned, for expressing the F_ENO, the exhalation flow rate (if indicated) was given as a subscript in mL·s⁻¹. For instance, a flow rate of 50 mL·s⁻¹ was written F_ENO50. If there was no direct or indirect evidence regarding the flow rate (mL·s⁻¹), we only used F_ENO.

^bSputum eosinophilia/elevated sputum eosinophils defined as sputum eosinophil count ≥2.5% (69) or ≥3% (70) or >3% (68) or no definition (3, 44, 45, 49, 55, 59, 61, 67, 71).

^cElevated blood eosinophils defined as blood eosinophil count ≥1% (69) or ≥200 eosinophils·μL⁻¹ (67) or no definition (45, 59).

^dSubjects/patients are COPD unless stated otherwise.

Optimal cut-off was presented in three studies (68–70), 17.5, 19, and 23.5, to identify sputum eosinophilia (elevated sputum eosinophils). The area under the receiver operating characteristic (ROC) curve (AUC) for the cut-off 17.5 and 19 ppb was 0.61 and 0.89, respectively. There was no report of AUC for the cut-off 23.5 ppb. The cut-off values of 17.5, 18, and 23.5 had a sensitivity of 64.5, 90 and 62.1%, and a specificity of 56.4, 74, and 70.5%, respectively. The study with the cut-off 23.5 showed that the F_ENO50 >23 ppb could be a good prediction of sputum eosinophilia (sputum eosinophil count >3%).

Fractional exhaled nitric oxide and treatment response

The studies (N = 8, 21.05%) included in this theme are presented in Table 6. Only one study was a RCT (49). Treatment response was defined as an increase in FEV1 > 12% and 200 mL in two studies (39, 70), ≥ 200 mL in one study (50), and >200 mL in another study (54). Three studies did not define the treatment response (3, 71, 72). Five of the eight studies reported an association between F_ENO50 levels in COPD and response to treatments, one study (50) with ICS, one with GCS administered either intravenous or orally (49) and three with the combination of inhaled ICS/bronchodilator (39, 54, 70). One study (3) reported an association with F_ENO50 levels in ACO with ICS therapy. According to this study (3), no significant difference reported between mild and moderate ACO with healthy subjects after 6 months ICS therapy while F_ENO50 increased in severe and extremely severe ACO. ACO and its severity were defined according to the GINA-GOLD 2014 (8) and GOLD 2011 report (GOLD stages ABCD) (73), respectively. Two studies (71, 72) reported no association between F_ENO50 levels and FEV1 pre- and post-bronchodilator therapy without ICS.

Table 6. Fractional exhaled nitric oxide and treatment response^a.

References	Treatment response influence/association^b	Optimal cut-off value	Comments^c
Kunisaki et al. (50)	Y	–	Higher baseline FENO50 levels in ICS responder than non-responders
Dummer et al. (49)	Y	50 ppb	Significant improvement in FEV1 from the lowest to the highest FENO50 tertile using oral glucocorticoid
Antus et al. (39)	Y	26.8 ppb	Higher increase in FEV1 and FEV1% predicted at discharge in the exacerbated patients with FENO50 > 26.8 ppb by administrating systemic (IV or orally) glucocorticoids and bronchodilators
Akamatsu et al. (54)	Y	35 ppb	Higher increase in FEV1 in patients with FENO50 >35 ppb and atopy by adding ICS/LABA
Soter et al. (70)	Y	26.8 ppb	Higher increase in FEV1 at admission in exacerbated patients with FENO50 >26.8 ppb by administrating systemic (IV or orally) glucocorticoids and bronchodilators
Amer et al. (72)	N	–	No association between FENO50 and change in FEV1 after bronchodilator therapy
Feng et al. (3)	Y	–	Association between FENO50 and ICS therapy (for 6 months) in ACO. Lower FENO50 in mild and moderate than severe and extremely severe ACO^d
Zhao et al. (71)	N	–	No association between the FENO50 and change in FEV1 after bronchodilator therapy

F_ENO: fractional exhaled nitric oxide; ICS: inhaled corticosteroid; ppb: parts per billion; FEV1: forced expiratory volume in first second; COPD: chronic obstructive pulmonary disease; IV: intravenous, ACO: asthma-COPD overlap; GINA: Global Initiative for Asthma; GOLD: Global Initiative for Chronic Obstructive Lung Disease.

^a Treatment response influence/association: Y = Yes, N = No, Optimal cut-off value: – = Not mentioned, for expressing the F_ENO, the exhalation flow rate (if indicated) was given as a subscript in mL·s⁻¹. For instance, a flow rate of 50 mL·s⁻¹ was written F_ENO50. If there was no direct or indirect evidence regarding the flow rate (mL·s⁻¹), we only used F_ENO.

^bTreatment response was defined as an increase in FEV1 ≥ 200mL (50) or >200mL (54) or >12% and >200 mL (39, 70). Treatment response was not defined in three studies (3, 71, 72).

^c Subjects/patients are COPD unless stated otherwise

^dACO and its severity was defined as GINA-GOLD joint document 2014 (8) and GOLD 2011 stages (ABCD) (73), respectively.

According to the five studies on COPD patients (39, 49, 50, 54, 70) included in this theme that showed an association with the levels of F_ENO50, patients who had the higher levels of F_ENO50 responded better to ICS/GCS with or without bronchodilator therapy than those who had the lower F_ENO50 levels. Four of the five studies introduced an optimal cut-off value for the treatment response (39, 49, 54, 70). Two studies reported the same cut-off value for the F_ENO50 level regarding treatment response, which was 26.8 ppb (39, 70). According to one of these two studies (39), the sensitivity and the specificity of this cut-off value were 74% and 75%, respectively, with the area under the ROC curve of 0.82. The other two studies (49, 54) proposed a cut-point of 50 and 35 ppb with a sensitivity of 29, 80% and a specificity of 96, 66.7%, respectively. The area under the ROC curve for the cut-off 50 ppb was reported as 0.69 (49).

Discussion

To the best of our knowledge, this is the first systematic scoping review undertaken on F_ENO and COPD. This scoping review started with a broad question: what do we know about F_ENO and its use in patients with COPD? The search criteria and inclusion criteria were deliberately kept broad in order to facilitate the inclusion of the largest number of studies possible. This approach enabled the determination of six themes and six specific questions. This, in turn, allowed us to conduct a review on the existing evidence and to identify gaps in the literature and make recommendations for clinical use and research.

The most extensive covered theme, mentioned in more than 60% of the articles, was the factors modifying F_ENO levels, followed by F_ENO and disease severity or progression in 57.89%, F_ENO in COPD compared with healthy subjects in 47.67%, F_ENO and biomarkers in 31.57%, F_ENO and treatment response in 21.05%, and F_ENO in ACO in 18.42%. Most of the included studies (N = 33) in this review mentioned that they measured F_ENO levels according to the ATS/ERS guideline 2005 (35). Five articles (40, 43, 47, 48, 63) did not mention that they used ATS/ERS guideline 2005 (35) to measure F_ENO.

We have seen from this review that when measuring F_ENO, there is a need to account for some important factors that could modify the level of F_ENO. Current cigarette smoking which is often present in COPD decreases F_ENO/F_ENO50 level (39–46) while COPD exacerbations increase F_ENO/F_ENO50 level (41, 51, 58). Concerning ICS/GCS alone or combined with bronchodilators, studies yielded conflicting results. On the one hand, seven studies (3, 39, 49–51, 54, 55) showed a decrease of F_ENO/F_ENO50 level with ICS/GCS while six studies (41, 43, 46, 48, 52, 53) showed no association. The difference in these results may be due to studies being underpowered, dealing with different COPD phenotypes (acute exacerbations, stable COPD, and ACO), different measurement method or use of different exhalation flow rate, and different devices being used for the measurement of F_ENO. Many factors have not been studied enough, preventing us from being able to make a definitive recommendation on those specific factors. Cold weather and/or viral infection might increase F_ENO level; sodium bicarbonate and exercise could decrease the F_ENO level. Of importance, we should also take into consideration that differences of measurement in the studies could be related to the type of device/instrument used for measuring F_ENO such as the chemiluminescence analyzer, electrochemical F_ENO device, NioxMino/Vero analyzer, SV-02 NO Instrument, and HypAir-FeNO analyzer. Using different devices might induce variability and limit comparison between studies. What we take from this review, before patients are tested for F_ENO, they should be advised not to smoke, to do exercise, and if possible to be off ICS or GCS. If the patient has an exacerbation and/or an acute respiratory infection such as a cold-like illness, F_ENO testing should be postponed. Moreover, we recommend conducting more studies, in particular on factors that have not been covered or those only covered by one or two studies.

Although F_ENO levels were generally higher in COPD compared with healthy individuals, none of the studies could propose a cut-off value to differentiate COPD from healthy subjects. Three studies (3, 4, 63) recommended conducting investigations with a large number of subjects to evaluate fractional exhaled nitric oxide measurement. According to Chen et al. (4), in order to generalize utilization of F_ENO50 measurement in the clinical setting, conducting prospective studies in large-scale would be crucial. According to Beg et al. (60), development of new handheld analyzers may make the measurement easier, make the large comparative studies possible, and lead to possible application in the daily clinical setting. Until this is demonstrated or refuted, F_ENO measurement is unlikely to be utilized in clinical use to differentiate COPD from non-COPD individuals.

Based on our review, F_ENO50 could be useful to differentiate phenotypes of COPD, more particularly ACO. All studies demonstrated that F_ENO50 has a higher level in ACO patients than those with only COPD (4, 25, 27, 53, 64, 65, 67). However, different cut-off values have been proposed for differentiating ACO from COPD which may be due to the use of different definitions of ACO as well as using different devices to measure F_ENO50. There is still uncertainty in identifying ACO among COPD patients, in particular, defining the optimal cut-off value to be used in clinical practice. Accordingly, Chen et al. (4) and Goto et al. (65) suggested establishing further prospective studies with a large number of subjects to investigate and broaden the utilization of F_ENO50 measurement in the clinical settings. Until we come up with a standard on which we can define and rely on a definition of ACO, it will be difficult to be definitive on a cut-off of F_ENO50 even with large sample size study.

The review also addresses the question of F_ENO and disease severity and progression. Studies reported an association with exacerbations, four out of six studies (42, 51, 52, 58), but not with stage of disease defined by GOLD as ABCD or as stages I–IV, 6/7 (4, 25, 27, 28, 43, 62). This relationship with exacerbations, F_ENO50 being elevated when patients have exacerbations may just be a consequence of the exacerbations. We still do not know if F_ENO could be useful to diagnose different types of exacerbations to help guide therapy. F_ENO/F_ENO50 does not appear to be associated with disease progression, defined by changes in pulmonary function tests (27, 39, 41–43, 49, 50, 54–56, 58, 59, 64).

Another area of major focus is the identification of biomarkers that would be indicative of asthma, in particular, high sputum eosinophils, IgE and more recently the possibility of some threshold of blood eosinophils. F_ENO can be measured via a fast, non-invasive, reproducible, and easy way in close to real time which would offer a significant advantage on measurement such as sputum induction. According to our scoping review, although there was a positive relationship between F_ENO50/F_ENO45 levels with sputum eosinophils (3, 44, 49, 55, 59, 61, 68–70) and serum IgE (3, 45, 55, 68), there was no unified cut-off of F_ENO50/F_ENO45 level. Studies used different definitions of high sputum eosinophilia. Finally, we cannot be definitive on the relationship between F_ENO50 level and blood eosinophils, two studies reported an association (45, 59) versus two studies showed no association (67, 69). In addition, the studies reported no association, did not clearly mention if they included both current and ex-smokers with COPD in their analysis that can be a potential factor decreasing correlation between blood eosinophilia and F_ENO.

Aligned with the role that F_ENO might play in monitoring airway inflammation in COPD, F_ENO could predict corticosteroid responsiveness. COPD patients who had higher levels of F_ENO50 at baseline (before treatment) demonstrated a better response to the medication therapy (ICS/GCS with or without bronchodilators) (39, 49, 50, 54, 70). However, studies used different cut-off levels of F_ENO50 from 26.8 (similar in two studies) (39, 70) to 35 (54) and 50 (49) ppb, for assessing treatment response. It is still unclear if F_ENO could be used and which cut-off will have the best yield to predict treatment response to ICS/GCS. Moreover, there would be a need to conduct further investigations with well-designed longitudinal studies to determine if F_ENO can reflect eosinophilic airway inflammation with enough precision to be used in a therapeutic decisional algorithm and/or as an alternative to sputum induction for guiding therapy.

This scoping review has the strength of selecting all the studies in the broad field of F_ENO in COPD. Themes and questions were determined based on the existing studies and not excluding studies of a specific topic. Furthermore, there was no limitation of language as the review included all languages. Our scoping review like the other ones was not free of limitations. One of the limitations of our study was excluding the conference/meeting and no peer-reviewed abstracts as well as removing review articles. We searched major databases for this scoping review. However, there are other databases that were not searched. We did a comprehensive search and only searched for the intersection of population and intervention; therefore, published negative results studies would have been retrieved by the search. However, there is always a likelihood that some studies with negative results were not included in the databases and in our study leading to the possibility of publication bias. This scoping review was limited to the year 2005 onwards as this was the year that the first ATS/ERS guideline regarding F_ENO measurement was published (35). Another limitation is related to the quality assessment of the studies included in the review. We did not perform quality assessment of the studies as it is usually not done in scoping review, according to the scoping review guidance (29) and consistent with scoping review of clinical topics (37, 38).

Conclusion

In conclusion, when measuring F_ENO, there are several factors that can modify its measurement. This needs to be considered in clinical setting and research. The evidence is still lacking preventing us from recommending the general use of F_ENO in clinical practice for COPD patients. Although F_ENO level is higher in ACO patients than COPD-only, it is still unclear if there is a F_ENO cut-off that can be used to make the diagnosis of ACO and/or to guide therapy with ICS/GCS in COPD patients. According to current evidence, F_ENO does reflect airway eosinophilia and responsiveness to ICS/GCS also in COPD. This should encourage researchers to conduct large clinical studies assessing the ability of F_ENO to find those subjects with COPD who gain most from ICS. As well, the focus of future research should be to determine if F_ENO could be part of a cascade of a therapeutic decisional algorithm and/or as an alternative to sputum induction for guiding COPD therapy.

Declaration of interest

Dr Bourbeau reports public grants from Canadian Institute of Health Research (CIHR), Canadian Respiratory Research Network(CRRN), Foundation of the McGill University Health Center (MUHC), industry grants all administered by the Research Institute of MUHC (from Aerocrine, AstraZeneca, Boehringer Ingelheim, Glaxosmithkline, Grifols, Novartis), personal fees for conferences AstraZeneca, Boehringer Ingelheim, Glaxosmithkline, Grifols, and Novartis, outside the submitted work; Dr Barrecheguren reports personal fees from Menarini, GlaxoSmithKline, Gebro pharma, Novartis, Grifols, outside the submitted work; Dr Mostafavi-Pour-Manshadi, Dr Naderi and, Dr Dehghan have no disclosure.

Acknowledgment

The authors would like to thank their librarian at the McGill University Health Centre, Alex Amar, for his help in preparing search strategy document and searching databases as well as providing updated database search.