The Efficacy and Safety of Autologous Blood Patch for Persistent Air Leaks: A Systematic Review and Meta-Analysis

Persistent air leaks (PALs) are associated with prolonged hospital stays, contamination and sustained infection of the pleural space, and significant morbidity. A fistulous tract between the alveoli and the pleural space is referred to as an alveolar-pleural fistula (APF), whereas a fistulous tract between the bronchiole and the pleural space is referred to as a bronchopleural fistula (BPF). There is no consensus on the treatment, and multiple modalities exist for the management of persistent air leak (PAL). Autologous blood patch (ABP) is a relatively safe and inexpensive method that has been used for many years for the treatment of PALs. We conducted an electronic database search between 08/24/2022 and 08/27/2022 in PubMed, Embase, and Cochrane using keywords. The following keywords were used: "Blood patch" OR "Autologous blood patch" AND "pleurodesis." Our study included all original studies with the prime focus on the etiology of PALs, clinical characteristics, procedural details of ABP, and outcomes of the proposed treatment. The primary outcomes that were the focus of our study were the time to seal the air leak, the time to remove the chest tube after air leak cessation, and the time to discharge from the hospital. To determine the safety of ABP, we also evaluated the procedural outcomes. Our findings suggest a statistically significant decrease in the time to air leak cessation when compared to the control group (mean difference of -3.75 {95% CI: -5.65 to -1.85; P=0.001}) with considerable heterogeneity of I2=85% and P=0.001. However, the difference was not statistically significant when a lower dose of ABP (50 mL) was compared to a higher dose (100 mL) (mean difference of 1.48 {95% CI: -0.07 to 3.02; P=0.06}) and considerable heterogeneity of I2=80% and P=0.03. There was no statistically significant difference in the time to discharge when compared to the control group (mean difference of -2.12 {95% CI: -4.83 to 0.59; P=0.13}) and considerable heterogeneity (I2=95% and P<0.001). When compared to the control group, ABP did not provide any statistically significant difference in the risk ratio for infection (1.18 {95% CI: 0.52 to 2.65; P=0.70} and moderate heterogeneity {I2=33% and P=0.20}), pain (1.18 {95% CI: 0.52 to 2.65; P=0.70} and moderate heterogeneity {I2=33% and P=0.20}), and fever (0.54 {95% CI: 0.27 to 1.10; P=0.09} and no heterogeneity {I2=0% and P=0.50}). Our study concludes that using ABP caused a statistically significant decrease in the time to air leak cessation when compared to the control group. However, the procedure does not provide a statistically significant difference in the time to discharge from the hospital when compared to conservative treatment. Similarly, there was no statistically significant difference in the risk ratio for complications such as infection, pain, and fever when compared to conservative management. More studies need to be conducted to fully understand the efficacy and safety of ABP in the management of PALs.


Introduction And Background
Air leaks occur when air from one cavity of the body enters another cavity, which usually does not contain air. Pneumothorax is the most common example of this phenomenon, in which air escapes from the bronchial tree or alveoli and enters the pleural space through a fistulous tract. Treatment usually consists of observation with spontaneous resolution, insertion of a chest tube with wall suction, and, in severe cases, surgical intervention. When a chest tube is placed, the air leak is identified by bubbling in the chest drainage system. Air leaks that persist for more than 5-7 days are considered persistent air leaks (PALs) [1]. Patients with fistulous tracts are at risk of significant morbidity and mortality due to ventilation/perfusion mismatch and pleural contamination with respiratory flora [2]. Depending on the initial cause of persistent air leaks, a variety of treatment options have been explored. These include prolonged chest tube drainage, chemical pleurodesis, autologous blood patch (ABP) pleurodesis, and temporary unilateral valve placements [1]. Pleurodesis using autologous blood patches has been used to treat persistent air leaks with variable results. In this study, we reviewed the literature on autologous blood patch pleurodesis for PAL and its efficacy in treating the condition.

Search Strategy and Study Selection
This study followed Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines for systematic reviews and meta-analyses, which do not require protocol registration [3]. An electronic database search was conducted for relevant studies published from 08/24/2022 to 08/27/2022 in PubMed, Embase, and Cochrane using keywords. We used these search terms in each database: "Blood patch" OR "Autologous blood patch" AND "pleurodesis." This search included all original studies (cohort, cross-sectional, and case-control studies) describing the etiology and duration of air leaks, clinical characteristics, procedures of autologous blood patch with outcome and complications associated with the intervention, and commentaries and case series with more than 10 patients. The exclusion criteria included non-original reports, which were either reviews, letters to editors, or commentaries that did not include patient data; case reports or case series of less than 10 patients; unextractable or irrelevant data; articles not published in English; duplicate records; animal studies; overlapped data; and full texts that were not available, unextractable, or irrelevant data.
The primary outcomes were the time to seal the air leak, the time to remove the chest tube after the air leak was stopped, and the time to discharge from the hospital following the air leak's cessation. A secondary outcome was the percentage of success and complications. In order to ensure that we did not miss any relevant studies, we manually searched the references of our included papers. All original studies that reported persistent air leaks and ABP were included in the study. Our systematic review was screened by two independent reviewers for titles and abstracts, followed by a full-text screening to ensure that relevant papers were included. We resolved disagreements by discussion and by referring them to the senior author when necessary.

Data Extraction
We developed a data extraction sheet using Microsoft Excel (Microsoft® Corp., Redmond, WA). Two independent reviewers extracted data using the Excel sheet. Disagreements and discrepancies were all resolved through discussions with the senior author.

Quality Assessment
The risk of bias in the included studies was evaluated by one independent reviewer. A risk-of-bias assessment tool developed by the National Institutes of Health (NIH) was used to assess the quality of the included studies [4].

Statistical Analysis
A descriptive analysis was conducted using Statistical Package for Social Sciences (SPSS) (IBM SPSS Statistics, Armonk, NY). Review Manager (RevMan) 5 (Cochrane, London, England) was used to develop the forest plot and funnel plot.

Search Results
Following the elimination of 30 duplicate records with EndNote 20 (Clarivate, London, England), we identified 101 records. Based on the title and abstract screening, 41 records were selected for further fulltext screening. After excluding 23 papers from the full-text screening phase, 18 studies were included in our study ( Figure 1).

Study Characteristics and Quality of the Included Studies
The baseline characteristics of the included studies are summarized in Table 1. There were 18 studies included: six were retrospective [5][6][7][8][9][10], seven were randomized controlled trials (RCTs) [11][12][13][14][15][16][17], one study was a non-randomized controlled trial [18], three were prospective studies [19][20][21], and one was a case series involving 11 patients with original data [22]. The sample sizes of the included studies ranged from 11 to 167 [22,20]. Patients ranging in age from adults to geriatrics were included in the study. Male patients constituted 77% of the study population.   [5]. Andreetti et al. report that adenocarcinoma is the most common cause of PAL in their study population, followed by squamous cell carcinoma [11]. Ferraroli et al. reported similar findings, with adenocarcinoma being the most common etiology of PAL [7]. Lang-Lazdunski et al. reported the most common pathology in patients with persistent air leaks to be adenocarcinoma, but lymphoma, large cell carcinoma, and Pancoast tumor have each been reported in one patient [22]. According to Akar et al. [6], Cao et al. [12], and Zhang et al. [13], secondary spontaneous pneumothorax is the primary cause of PAL.
In a few studies, primary and secondary pneumothorax have been identified as the etiology of PALs, but the underlying cause of secondary pneumothorax has not been described [14,16,20,21]. Interstitial lung disease (ILD) and tuberculosis were also suggested as the probable causes of PAL in some studies reported [6,9,10,17,19]. A minimum duration of air leaks of 2-11 days was required for study participants to be eligible for ABP. A median of seven days has been reported by Dye et al. [8], and the range is from four to nine days. A wide range of interventions was performed in the study subjects prior to ABP, the most common of which was tube thoracostomy (since the intervention was necessary to manage an air leak that caused a pneumothorax to develop).

Procedural Details
Apilioğulları et al. [5], Cao et al. [12], and Khan et al. [16] conducted ABP with 1-2 mL/kg, 0.5-2 mL/kg, and 1 mL/kg of blood, respectively. Andreetti et al. reported a study design where 50 mL of autologous blood was infused into group A participants and 100 mL of blood was infused into group B participants, with group C serving as an observational group [11]. In a similar study design, Akar et al. infused 60 mL of blood into participants in group A and 120 mL into participants in group B [6]. Ferraroli et al. [7], Zhang et al. [13], Shackcloth et al. [15], and Martínez-Escobar et al. [18] performed studies in which the experimental groups received 10 mL, 20-30 mL, 120 mL, and 50-75 mL of autologous blood, respectively. Dye et al. conducted a retrospective study in which subjects receiving lung mass resection received 45-120 mL and 140 mL in their first and second attempts. During the first and second attempts, the lung volume reduction group received 75-120 mL and 80-110 mL, respectively. In the remaining studies, patients in experimental groups received 50 mL of autologous blood [8].

Outcomes, Complications, and Results of the Meta-Analysis
Our meta-analysis included eight studies [6,11,[13][14][15][16]18,20]. This meta-analysis focused on the time to seal air leaks (days) and the time to discharge from the hospital (days). This meta-analysis focused on complications related to fever, infection, and pain. The time to seal an air leak (days) was compared between the experimental group receiving 50 mL and the control group.

Risk of Bias of Included Studies
Based on the NIH tool, all of the studies included scored at least a 10 on the assessment. We evaluated the risk of bias of the case series included in our study using the NIH risk assessment tool ( Table 2) [4].   (9) and two days (4); control group, one day (7) and two days (6) --

Discussion
PAL refers to the detection of an air leak for a period exceeding 5-7 days [1]. A fistulous tract causing persistent air leak may be located between the alveoli or bronchioles and the pleural space and is termed an alveolar-pleural fistula (APF) or a bronchopleural fistula (BPF). PAL is most commonly caused by spontaneous pneumothorax, with or without underlying lung disease [1,2]. Among the other causes are pulmonary infections, trauma, thoracic surgery involving lung parenchyma, and mechanical ventilation complications [1].
The quantification of PAL plays an important role in the management of PAL. A three-chamber drainage system is a useful tool for identifying and quantifying air leaks. Recently, Chambers et al. developed a fourgrade classification system to determine the severity of air leaks after thoracic surgery [1,2,23,24]. The most severe form of PAL is continuous (C) air leak, which occurs throughout the respiratory cycle. A patient with this type of air leak would require mechanical ventilation with a large underlying BPF [1]. An inspiratory (I) air leak is similarly rare and would require a large fistulous defect in order to occur. Expiratory (E) air leaks are defined as leaks discovered during expiration. The fourth type of leak is found only during forced expiration (FE) or coughing and is known as a forced expiration air leak. However, this system of classification is generally better suited to patients who have undergone thoracic surgery and is not generally applicable to patients who are critically ill [25]. The decision regarding treatment in such patients is usually based on the clinical condition of the patient.
The management of PAL is usually determined by the severity of the treatment and the underlying etiology. The initial assessment is made by sequential balloon inflation and occlusion or injections of methylene blue in order to locate the defect. There are several treatment options available, ranging from conservative management through extended chest tube drainage to chemical and autologous blood patch pleurodesis, endobronchial valve placement, and surgical corrections via video-assisted thoracoscopic surgery (VATS) or open thoracotomy involving mechanical or chemical pleurodesis or pleurectomy [1,2].
A variety of studies have been conducted on autologous blood patches, and they were previously used as a preventative measure for recurrent spontaneous pneumothorax. However, the literature now indicates that it is a safe, effective, and affordable method of treating persistent air leaks caused by pneumothorax [1,2,[26][27][28][29][30][31][32][33][34][35][36][37]. It is believed that autologous blood patches work in two ways. Firstly, the air leak is directly sealed by a blood clot, alongside pleural inflammation caused by blood products. Secondly, the blood clot physically occupies the pleural space, thereby reducing leakage [2].
Ferraroli et al. report a primary study comparing intrapleural fibrin glue with autologous blood patch to surgical thoracotomy. In the study, no statistical significance was found between the two methods, suggesting that a conservative approach with autologous blood patches might reduce the need for surgical intervention and the complications associated with it [7].
According to Chambers et al., they reviewed 10 studies and found that the combined success rate of autologous blood patches is 93%, with 70%-81% of leaks being resolved within 12 hours and 95%-100% within 48 hours [2,30]. In contrast, conventional treatment with chest tube drainage requires 3-6 days of simple tube thoracostomy management [2].
This procedure consists of instilling 50-100 mL of autologous blood through the chest wall into the patient's pleural cavity under sterile conditions [1,2]. Following the flushing and clamping of the tube for around 30-60 minutes, the pleural space is suctioned. Among the most concerning complications of the procedure are tension pneumothorax and infections [2]. It is important to follow up closely with patients before and after the procedure to ensure they are hemodynamically stable and that there is no sign of infection.
Based on the studies included in our review, it appears that ABP is beneficial in the setting of PAL. Ibrahim et al. reported a mean time to seal PAL to be five days as opposed to 10 days in the observation group [14]. Lillegard et al. found that five out of eight patients improved almost instantly, with one patient improving on day 1 and two patients improving on day 2 [28]. There does not seem to be an obvious relation with the dose of autologous blood administered, as evidenced by a study reported by Cao et al., which reports a rather unpredictable pattern to improvement in PAL after the administration of different doses of autologous blood patches [12].
A meta-analysis was conducted during this study in order to evaluate the clinical outcomes and complications associated with the treatment of PAL with autologous blood patches. During the metaanalysis, we observed a clinically significant reduction in air leak repair times in patients who received autologous blood patches, as compared with our control group, as reported by Andreetti et al. [11], Ibrahim et al. [14], and Zhang et al. [13] (Figure 2 and Figure 3). Furthermore, there was no clinical difference between 50 mL of autologous blood and 100 mL of blood used during the procedure, as reported in the meta-analysis of the studies reported by Akar et al. [6] and Andreetti et al. [11] (Figure 4 and Figure 5).
We noted that patients who received the treatment with autologous blood patch did not show any statistically significant decrease in the time to discharge as opposed to the control group, which was reported in the meta-analysis of the studies of Cagirici et al. [20], Ibrahim et al. [14], and Zhang et al. [13] ( Figure 6 and Figure 7). Data for infection, fever, and pain was evaluated as complications of the treatment with autologous blood patch. The meta-analysis performed on the available data shows no statistical significance between the experimental and control arm in terms of infection rates and pain ( Figure 8 and Figure 9); however, it does show significance in terms of fever in the experimental group ( Figure 8 and Figure 9) [11][12][13][14][15][16].
A literature review also shows an improvement in PAL in children, as well as improved clinical outcomes [28,29]. Current evidence indicates that ABP may be considered a gold standard or first-line treatment for PAL caused by certain medical conditions, such as acute respiratory distress syndrome (ARDS) and ILD [37]. However, a lack of RCT data may be viewed as a potential disadvantage. It is necessary to collect more data through RCTs in order to quantify the potential magnitude of benefits associated with ABP.

Conclusions
Recent data regarding the treatment of persistent air leaks with autologous blood patches indicates promising results that suggest that the conservative management of persistent air leaks with autologous blood patches may be comparable in treatment efficacy and outcome without surgical intervention. Further research is required, however, in order to fully assess the non-inferiority of autologous blood patches in comparison with surgical treatment and other conventional treatments.

Conflicts of interest:
In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work.