ISSN : 1301-5680

e-ISSN : 2149-8156

Preoperative risk factors of airway complications in adult lung transplant recipients: A systematic review and meta-analysis

Mahmut Subasi^{1}, Mustafa Duger^{2}

**Methods: **Articles published between November 1995 and
February 2023 were searched by a thorough exploration of
databases. Studies that addressed recipient-related risk factors
for airway complications following adult lung transplantation,
such as cohorts, case-control, or cross-sectional studies, were
included. Fixed-effects or random-effects models were used to
calculate the odds ratios (ORs) or mean differences (MDs) with
95% confidence interval (CI).

**Results: **Twenty-one studies including a total of 38,321 recipients
fulfilled the inclusion criteria. Based on the pooled analyses,
taller height (MD=5.98, 95% CI: 5.69-6.27, I2 = 57.32%),
intraoperative mechanical ventilation (OR=1.83, 95% CI:
1.41-2.38, I2=0%), male sex (OR=1.52, 95% CI: 1.33-1.74,
I2=15.91%), preoperative extracorporeal membrane oxygenation
(OR=1.58, 95% CI: 1.1-2.26, I2 =41.47 %), and preoperative steroid
use (OR=1.21, 95% CI: 1.04-1.41, I2 = 0%) elevated the risk of
airway complications following lung transplantation.

**Conclusion: **Taller height, intraoperative mechanical ventilation,
male sex, preoperative extracorporeal membrane oxygenation,
and preoperative steroid use can increase the risk of airway
complications after lung transplantation. Identifying high-risk
recipients or riskless situations can support the advancement
of selective treatments or prevent the unnecessary avoidance of
certain interventions.

Lung transplantation (LTx) is the most effective
form of treatment for end-stage lung disease. It also
increases survival and quality of life. Despite the
extensive surgical and postoperative management
advances in this field, airway complications
(ACs) remain a common cause of morbidity and
mortality.[1] There are many different types of AC
such as granulation, stenosis, tracheobronchomalacia,
bronchial fistula, anastomotic infection, and
dehiscence. A significant contributing factor to AC
has been found as postoperative decreased blood flow
in the donor bronchus, despite the fact that the exact
mechanism is yet unknown.[2]

Lung transplantation is the only solid organ transplantation in which systemic arterial blood supply is not routinely anastomosed.[3] Traditionally, bronchial arteries have been severed during transplantation. This results in ischemia until pulmonary artery collaterals develop in the submucosal plexus; this frequently takes weeks or months to establish. By boosting arterial resistance and interfering with collateral development, postoperative interstitial edema and reperfusion injury may be more responsible for anastomotic ischemia consequences.[1] As a result, the bronchial anastomosis is allowed to repair while being ischemic.[4]

A variety of potential risk factors for the development of AC have been found. These risk factors can be associated with the donor or recipient, surgical techniques, infections, medications, or immunosuppression.[5] Numerous research have been conducted to identify these risk variables.[2,6-25] In this systematic review and meta-analysis, we identify the most significant preoperative risk factors for ACs among adult LTx recipients.

The PubMed and ISI Web of Science were systematically searched for articles published between November 1995 and February 2023. The following keywords were used in searching: "Airway Complications" and "Lung Transplantation". There were no limits on languages. The literature search was independently assessed based on the title, abstract, or descriptors to locate possibly pertinent papers for in-depth assessment. References from primary or review papers were also manually checked to look for any more applicable studies.

The meta-analysis included original studies comparing groups with and without ACs in terms of risk factors. Studies without a control group, those concentrating on pediatric LTx or re-LTx, or those managing ACs were not included. The rat or animal experiments were also excluded. Case reports, case series studies, image interests, comments, and full texts that could not be accessed were not considered. Studies from the same author, study group, or institution that were the longest or most recent series were included, while others were excluded.

**Selection criteria**

Cohort, case-control, and cross-sectional studies
were included, if they investigated which recipient"s
factors directly influencing the development of ACs
after LTx. Variables included age, male sex, body mass
index (BMI), height, ischemic cardiac disease, diabetes
mellitus (DM), preoperative diagnoses (i.e., chronic
obstructive pulmonary disease [COPD], cystic fibrosis,
pulmonary fibrosis, pulmonary hypertension), prior
thoracic surgery, cytomegalovirus (CMV) positivity,
microbiological colonization, preoperative steroid use,
intraoperative mechanical ventilation, and preoperative
extracorporeal membrane oxygenation (ECMO). After
obtaining the complete text of the papers, the authors
separately assessed eligibility. After the differences
between the two reviewers were resolved, they were
able to agree on the final set of data by reviewing
relevant papers.

**Data extraction**

Two researchers independently compiled
summaries of the papers that met the inclusion
criteria and extracted data using a common data
sheet. The following data were extracted from each
study: first author's name, study design, publication
year, study date (initial and end), country, comparison
groups, sample size, number of postoperative ACs,
AC delineations, the Newcastle-Ottawa Scale (NOS)
(Table **1**).

**Table 1: Baseline characteristics of the included studies**

**Study quality evaluation**

Based on the following nine questions, the NOS
was used to rate the excellence of observational
studies: The following criteria must be met:
(i) representativeness of the exposed cohort; (ii) choice
of the non-exposed cohort; (iii) determination
of exposure; (iv) proof that the outcome was not
present at the start of the study; (v) comparability;
(vi) assessment of outcome; (vii) a dequate l ength o f
follow-up; (viii) adequate participant follow-up; and
(ix) total stars. There is a maximum score of 9 on this
scale. A total score of 7 to 9 was considered "good,"
a score of 4-6 was considered "fair," and a score of
4 was considered "poor."

**Statistical analysis**

Statistical analysis was performed using the
Medical Research Support (MedicReS E-PICOS
Version 21.3, NY, United States) program. The
odds ratios (ORs) and 95% confidence intervals
(CIs) were calculated to estimate the association
between binary factors (age, male sex, BMI, height,
ischemic cardiac disease, DM, COPD, cystic fibrosis,
pulmonary fibrosis, pulmonary hypertension, prior
thoracic surgery, CMV positivity, microbiological
colonization, preoperative steroid use, intraoperative
mechanical ventilation, and preoperative ECMO) and
development of AC. We assessed the mean differences
(MDs) for LTx subjects with and without AC, when
the mean values and SDs for a specific risk factor
were given. Fixed-effects models or random-effects
models were used to produce the statistical estimates
of effect based on I2. The I2 statistic was used to
quantify heterogeneity. Using accepted guidelines,
an I2 between 0 and 40% was considered to exclude
heterogeneity, 30 and 60% moderate heterogeneity,
50 and 90% substantial heterogeneity, and 75 and
100% considerable heterogeneity. Publication bias
was assessed with funnel plots.

**Figure 1: Flowchart of the meta-analysis.
* 216 papers on Web of Science and 218 on PubMed.**

According to the NOS, all studies had excellent
methodological quality (good or fair) (Table **1**).

Age: S eventeen s tudies ( sample s ize=37,588) examined the impact of recipient age on the happening of AC following LTx.[2,6-11,13,15-21,24,25] The outcomes of this analysis displayed no significant difference in the mean age between patients who had AC (n=987) and those who did not (n=36,601) (MD=0.56, 95% CI: -0.72-1.84, p= 0.69). Heterogeneity was considerable (I2 = 95.81%, p<0.001) and random model and meta regression were used.

Male sex: N ineteen s tudies ( sample s ize=38,216)
examined the impact of male sex on the happening
of AC following LTx.[2,6-11,13-20,22-25] The outcomes of
this analysis presented a significant difference in
in the proportion of male sex between patients who
had AC (n=1,112) and those who did not (n=37,104)
(OR=1.52, 95% CI: 1.33-1.74, p<0.001). Populations
were homogeneous (I2 = 15.91%, p=0.26) and fixed
effect model was used (Figure **2**).

**BMI:** Seven studies (sample size=17,681) examined
the impact of recipient BMI on the happening of
AC following LTx.[2,7-10,13,25] The outcomes of this
analysis exhibited no significant difference in the
mean BMI between patients who had AC (n=434)
or and those who did not (n=17,247) (MD=0.72, 95% CI: -0.73-2.18, p = 0.33). Heterogeneity was
considerable (I2 = 95.1%, p<0.001 ) and random model
and meta regression were used.

**Height:** Three studies (sample size=674) examined
the impact of recipient height on the happening of AC
following LTx.[9,18,20] The outcomes of this analysis revealed a significant difference in the mean height
between patients who had AC (n=75) or those who did
not (n=599) (MD=5.98, 95% CI: 5.69-6.27, p<0.001).
Heterogeneity was moderate (I2 = 57.32%, p=0.1) and
fixed model was used (Figure **3**).

Ischemic cardiac disease: Four studies (sample size=854) examined the impact of ischemic cardiac disease on the happening of AC following LTx.[6,9,16,20] The outcomes of this analysis indicated no significant difference in the proportion of ischemic cardiac disease between patients who had AC (n=141) and those who did not (n=713) (OR=0.74, 95% CI: 0.46-1.16, p=0.19). Heterogeneity was moderate (I2 = 30.27%, p=0.24) and fixed effect model was used.

DM: Five studies (sample size=34,591) examined the impact of DM on the happening of AC following LTx.[2,6,9,11,13] The outcomes of this analysis showed no significant difference in the proportion of DM between patients who had AC (n=588) and those who did not (n=34,003) (OR=1.06, 95% CI: 0.86-1.31, p=0.6). Populations were homogeneous (I2 = 0%, p=0.52). Fixed effect model was used.

**COPD: **Fourteen studies (sample size=19,138)
examined the impact of COPD on the happening of
AC following LTx.[6,8-10,13,16-23,25] The outcomes of this
analysis displayed no significant difference in the
proportion of COPD between patients who had AC
(n=645) or and those who did not (n=18,493) (OR=0.91,
95% CI: 0.76-1.1, p=0.33). Heterogeneity was moderate
(I2 =37.18 %, p=0.08) and fixed effect model was used.

**Cystic fibrosis:** Ten studies (sample size=18,242)
examined the impact of cystic fibrosis on the
happening of AC following LTx.[9,10,13,14,16-21] The
outcomes of this analysis presented no significant
difference in the proportion of cystic fibrosis between
patients who had AC (n=583) and those who did not
(n=17,659) (OR=0.99, 95% CI: 0.78-1.25, p=0.92).
Study populations were homogeneous (I2 = 9.08%,
p=0.36) and fixed effect model was used.

**Pulmonary fibrosis:** Sixteen studies (sample
size=19,863) examined the impact of pulmonary
fibrosis on the happening of AC following
LTx.[2,6-9,13-23] The outcomes of this analysis
displayed no significant difference in the proportion of pulmonary fibrosis between patients who had
AC (n=758) or and those who did not (n=19,105)
(OR=1.06, 95% CI: 0.9-1.26, p=0.49). Heterogeneity
was moderate (I2 =47.11%, p=0.02) and fixed effect
model was used.

**Pulmonary hypertension:** Ten studies (sample
size=19,013) examined the impact of pulmonary
hypertension on the happening of AC following
LTx.[2,6,8,10,13,14,17,18,20,23] The outcomes of this analysis
showed no significant difference in the proportion
of pulmonary hypertension between patients who
had AC (n=649) and those who did not (n=18,364)
(OR=1.04, 95% CI: 0.74-1.47, p=0.82). Heterogeneity
was moderate (I2 = 34.2%, p=0.14) and fixed effect
model was used.

**Prior thoracic surgery: **Three studies (sample
size=18,378) examined the impact of prior thoracic
surgery on the happening of AC following LTx.[2,8,11]
The outcomes of this analysis indicated no significant
difference in the proportion of prior thoracic surgery
between patients who had AC (n=315) and those
who did not (n=18,063) (OR=1.13, 95% CI: 0.77-1.65,
p=0.54). Study populations were homogeneous (I2 = 0%,
p=0.65) and fixed effect model was used.

**Cytomegalovirus positivity:** Three studies
(sample size=757) examined the impact of CMV
positivity on the happening of AC following
LTx.[16,18,20] The outcomes of this a nalysis exhibited
no significant difference in the proportion of CMV
positivity between patients who had AC (n=99) and those who did not (n=658) (OR=0.95, 95%
CI: 0.62-1.47, p=0.83). Study populations were
homogeneous (I2 = 0%, p=0.98) and fixed effect model
was used.

**Microbiological colonization: **Six studies
examined (sample size=1,247) the impact of
microbiological colonization on the happening of AC
following LTx.[2,7,9,14,19,22] The outcomes of this analysis
displayed no significant difference in the proportion
of microbiological colonization between patients who
had AC (n=224) and those who did not (n=1,023)
(OR=0.87, 95% CI: 0.4-1.89, p=0.72). Heterogeneity
was considerable (I2 = 82.47%, p<0.001) and random
effect model was used.

**Preoperative steroid use:** Eleven studies (sample
size=36,038) examined the impact of preoperative
steroid use on the happening of AC following
LTx.[2,8,11-13,18-23] The outcomes of this analysis
revealed a significant difference in the proportion of preoperative steroid use between patients who had AC
(n=734) and those who did not (n=35,304) (OR=1.21,
95% CI: 1.04-1.41, p=0.02). Study populations were
homogeneous (I2 = 0%, p=0.88) and fixed effect model
was used (Figure **4**).

**Intraoperative mechanical ventilation: **Six
studies (sample size=34,875) examined the impact of
intraoperative mechanical ventilation on the happening
of AC following LTx.[2,6,11,13,19,21] The outcomes of this
analysis revealed a significant difference in the
proportion of intraoperative mechanical ventilation
between patients who had AC (n=613) and those
who did not (n=34,262) (OR=1.83, 95% CI: 1.41-2.38,
p<0.001). Study populations were homogeneous
(I2 =0 %, p=0.82) and fixed effect model was used
(Figure **5**).

**Preoperative ECMO:** Six studies (sample
size=35,404) examined the impact of preoperative
ECMO on the happening of AC following LTx.[2,8,9,11,13,24] The outcomes of this analysis
revealed a significant difference in the proportion
of preoperative ECMO between patients who had
AC (n=592) and those who did not (n=34,812)
(OR=1.58, 95% CI: 1.1-2.26, p=0.01). Heterogeneity
was moderate (I2 =41.47 %, p=0.14) and fixed effect
model was used (Figure **6**). The forest plot of the all
parameters can be seen in Figure **7**.

Airway complications have been a major factor limiting the development of LTx throughout history and associated with considerable morbidity and mortality.[2] Among the risk factors of AC, male sex was a significant risk factor in our analysis. The bronchial arteries' origins, number, dimensions, and courses can differ greatly among individuals and between sexes. [27] Men have much more bronchial arteries than women, both in terms of size and number. Men may, therefore, have lower ischemia tolerance.

A similar argument could be possible for the height of the patients. The bronchus is greater in diameter in tall patients, and emphysema patients typically have less peribronchial fatty tissue, which can be used to cover the anastomosis.[20] This is probably due to a recipient-donor size discrepancy, as seen by the recipient's wider bronchial circumference and the donor bronchus' requirement for intussusception.[28] There will, therefore, be a requirement for telescopic anastomosis, which has the potential to result in more AC than end-to-end anastomosis. Similarly, our study supports that the strongest recipient's risk factor is the height.

Mechanical ventilation is an important component in the perioperative management of LTx. Mechanical ventilation poses a risk of bronchial ischemia, as it damages the bronchial mucosa and increases arterial resistance.[29] Ischemia makes the bronchial anastomoses sites more vulnerable to poor healing, infection, and complications with the anastomotic airway. Additionally, perfusion of transplanted airways may be compromised by positive pressure mechanical ventilation, particularly when large inflation pressures are necessary. The pulmonary flow to the main bronchi would be decreased by any allograft parenchymal pathology, such as primary graft disfunction, infection, or rejection, which would impair anastomotic recovery. Anastomotic stress and bronchial wall deterioration can be also caused by positive pressure ventilation. Some studies have found a connection between the likelihood of anastomotic ACs and high airway pressures and longer ventilation periods.[30] Our study showed that intraoperative mechanical ventilation increases the risk of AC after LTx.

Due to the interruption of microcirculation, inflammatory responses that cause endothelial damage, or issues specifically associated with ECMO, it is possible that ECMO would impair the healing of the airways. Due to the fact that ECMO is used to treat AC-related high-risk disorders such as primary graft disfunction, it can also seem to increase the risk.[2] In our study, there was also a slightly higher risk of ECMO.

It is well known that perioperative steroid medication has a negative impact on the repair of bronchial anastomoses; however, recent findings have indicated that this is debatable. High doses of steroids within the first year following surgery increase the chance of AC. Patients receiving high doses of corticosteroids run the risk of having worse early postoperative outcomes, which cannot be ruled out. To optimize the early preoperative course, it is crucial to customize and limit the preoperative steroid dose.[12] In our study, preoperative steroid use moderately increased AC.

Nonetheless, there are some limitations to this analysis. Although the evaluation or diagnosis of AC was made according to global standards, there was heterogeneity between studies. The follow-up period was also variable between studies, although it was long enough to yield results. Additionally, numerous problems were identified in detail in some research, while others only provided a single definition for them. The primary diseases leading to LTx, for instance, were noticed to be classified differently in each study. Our analysis covered the most frequently studied ones, and we concluded that the primary diagnosis indicative of LTx did not increase the risk of airway problems. Some of the studies also focus on a single variable, treatments or survival. The other limitations could be potential publication bias, heterogeneity, not all variables are comparable, cannot overcome subjectivity, and only deals with main effects.

In conclusion, our analysis show that recipient"s preoperative risk factors such as taller height, preoperative mechanical ventilation, male sex, preoperative extracorporeal membrane oxygenation, and preoperative steroid use can increase the risk of airway complications after lung transplantation based on pooled analyses. Identifying the high-risk recipients or riskless situations can support the personalized approaches such as advancement of selective treatments or prevent the unnecessary avoidances.

**Ethics Committee Approval: **The study protocol was
approved by the Istanbul Medipol University Non-Interventional
Clinical Research Ethics Committee (date: 28.07.2023, no:
E-10840098-772.02-4639). The study was conducted in
accordance with the principles of the Declaration of Helsinki.

**Data Sharing Statement: **The data that support the findings
of this study are available from the corresponding author upon
reasonable request.

**Author Contributions:** All authors contributed equally to
the article.

**Conflict of Interest:** The authors declared no conflicts of
interest with respect to the authorship and/or publication of this
article.

**Funding: **The authors received no financial support for the
research and/or authorship of this article.