Use of endo-bronchial end-tidal CO2 test for location of the pleural air leakage in patients with intractable pneumothorax.

BACKGROUND: Location of the affected bronchus of pleural air leaks is the most important step of trans-bronchoscopic bronchial occlusion for the treatment of intractable pneumothorax. The balloon occlusion test is the most commonly used technique, but has failed in some cases. The aim of the present study was: (1) to determine if endo-bronchial end-tidal CO2 (EtCO2) measurement can identify the affected bronchus that is the source of a persistent pleural air leak; and (2) to establish a methodology for endo-bronchial EtCO2 testing in locating affected bronchus in intractable pneumothorax.
METHODS: A total of 28 patients with intractable pneumothorax underwent bronchoscopy with (1) the balloon occlusion test for the identification of the affected bronchus; and (2) endo-bronchial EtCO2 measurement (EtCO2 test) at the orifices of the bronchus of the affected lung. The effectiveness of these two methods of affected bronchus identification were compared. The threshold EtCO2 (T-EtCO2) was determined.
RESULTS: The positive rates of locating the affected bronchus by the endo-bronchial EtCO2 test, balloon occlusion test, and combination of the two techniques were 60.7% (17/28), 64.3% (18/28) and 96.4% (27/28), respectively. The average differences in EtCO2 between the affected bronchus and the main carina, main bronchus, and non-affected bronchus were (in mmHg) 4.41 ± 1.99 (95% confidence interval: 3.5, 5.3), 4.73 ± 2.10 (3.80, 5.66 ) and 5.57 ± 2.53 (4.45, 6.69), respectively.
CONCLUSIONS: (1) The endo-bronchial EtCO2 test is complementary to the balloon occlusion test of the leading bronchus. (2) A threshold (T-EtCO2) value of >5 mmHg is optimal for this technique.

Introduction

Intractable pneumothorax or prolonged pulmonary air leak is defined as persistent pleural air leak for more than 7 days. Common causes of intractable pneumothorax are COPD, lobectomy and trauma. The bronchial occlusion technique has been proven to be an effective method for treating intractable pneumothorax.[1-4] One of the key steps in bronchial occlusion is to detect the bronchus leading to pleural air leakage – that is, the affected bronchus. In previous clinical research, the balloon occlusion test has been the most frequently used method for detecting the affected bronchus. In our previous research, the detection power of this technique was found to be 85%.[4] However, in subsequent clinical practice, we found that the balloon occlusion test failed to locate the affected bronchus in pneumothorax patients with multiple pleural leakages or collateral ventilation. The aim of the present study was to explore the value of endo-bronchial end-tidal CO2 (EtCO2) measurement in the detection of the affected bronchus, and furthermore to establish the exact methodology of trans-bronchoscopic EtCO2 detection.

Methods

A prospective observational study was conducted to evaluate the value of endo-bronchial EtCO2 test in identifying the affected bronchus. EtCO2 values at different anatomic sites of the tracheobronchial tree were checked, and the occlusion effectiveness of the combination of EtCO2 test and balloon occlusion test was also evaluated.

Recruitment criteria

Patients with persistent pleural air leaks from intercostal drainage for more than 7 days were included.

Exclusion criteria: (1) Patients were unable to tolerate the bronchoscopic procedure due to poor cardiopulmonary function. (2) Patients refused to undergo the bronchoscopic procedure. (3) Any other contraindications to bronchoscopy.

This research has been approved by the Institutional Review Board of the Second Affiliated Hospital of Fujian Medical University (approval number 2012-022). All procedures were performed in the same hospital and by the same bronchoscopists. Eligible patients were informed about the present study, and written informed consent was obtained from each participant. Participants were all from the Department of Pulmonary and Critical Care Medicine of the Second Affiliated Hospital of Fujian Medical University, treated between July 2012 and September 2015. There were 42 pneumothorax patients with intercostal drainage for more than 7 days. Of these, 31 patients agreed to participate in the present study, 3 of whom quit due to lack of tolerance of the bronchoscopic procedure; the remaining 28 patients completed the study.

Instruments

An Olympus CV-260SL flexible bronchoscope, MTN-SRB 3-lumen balloon catheter (Micro-Tech, Nanjing, China), and an EtCO2 detector (Microstream®, Oridion, Needham, MA, USA; Figure 1) were employed during the procedures.

Figure 1.

Instruments for the endo-bronchial EtCO2 test.

Determination of the affected bronchus

Flexible bronchoscopy (⩾2.8 mm i.d.) was performed in all patients under local anesthesia with 2% lidocaine and mild sedation (midazolam). Both the balloon occlusion test and EtCO2 detection were used to detect the affected bronchus according to the following procedures. For balloon detection, a balloon catheter was introduced through the working channel of a flexible bronchoscope, and the affected bronchus was detected as described in our previous publication.[4] Briefly, the balloon was inflated to achieve complete occlusion in the lobar, segmental and sub-segmental bronchus. The affected bronchus was identified by reduction or elimination of the air leak through the chest tube 15–20 s after occlusion. The EtCO2 test was performed as follows: an EtCO2 sampling catheter was introduced through the working channel of a flexible bronchoscope. The proximal end of the catheter was connected to an EtCO2 detector (Microstream®), and the exhaled gas was sampled from the distal end of the catheter with the help of the air pump inside the Microstream® (Figure 1). The EtCO2 was sampled at different anatomical sites (Figure 2): the lower segment of the trachea (above the main carina), the main bronchus of the affected lung, and the bronchial orifices for each lobe of the affected lung. At each site, the EtCO2 values usually stabilized within 10 cycles of breathing; subsequently, EtCO2 values were recorded for five cycles of relaxed breathing, and the average value at each site was calculated and recorded. A bronchus was considered as the suspected affected bronchus if its EtCO2 value was significantly lower than values from the other sites; further measurements were performed at the constituent segmental bronchial orifices if necessary.

Figure 2.

Example of the endo-bronchial EtCO2 test. The arrow points to the tip of the EtCO2 sampling catheter. Sampling sites: carina (A), main bronchus of left lobe (B), orifice of left upper lobe (C), orifice of left basal segments (D).

Occlusion of the affected bronchus

Once the affected bronchus was located, bronchial occlusion was performed to stop pleural air leakage. Materials for bronchial occlusion include either or both of autologous blood plus thrombin and endo-bronchial spigot. Technical details are offered in the literature.[3-5] When injected into the affected bronchus, autologous blood can coagulate in the sub-segmental bronchus in a short time, and the air leak can be stopped. The advantages of autologous blood are as follows: it does not cause immunoreactions in one’s own body; the blood clots can stop the air leak; and the blood clots can be absorbed within 1–2 weeks, in which time the bronchopleural fistula may be cured. Thrombin can accelerate the coagulation of autologous blood. Autologous blood plus thrombin for treatment of intractable pneumothorax has been proven to be an effective treatment choice in the clinic.[4,5]

Criteria for positive EtCO2 test

The EtCO2 test was ultimately defined as positive when the EtCO2 value at the orifice of one or more specific bronchus was lower than the values at other bronchial orifices, and the following criteria were met: (1) the result was consistent with the positive result of the balloon occlusion test; (2) the pleural fistula was successfully sealed after bronchial occlusion guided by the result of a reduction in EtCO2 value.

Determination of the reference point and the reference EtCO2 value (R-EtCO2)

After a specific bronchus (affected bronchus) was judged to be positive according to the positive criteria of the EtCO2 test as described above, its EtCO2 value was compared with EtCO2 values from the main carina, main bronchus, and other bronchus of the affected lung. When the EtCO2 value at any one of the three sites stated above was statistically higher than the EtCO2 value at the affected bronchus, the site with statistically higher value was determined to be the reference point, and the EtCO2 value at this site was labeled the reference EtCO2 value (R-EtCO2 value).

Establishment of the threshold value of EtCO2 (T-EtCO2) test

The difference of EtCO2 value between the R-EtCO2 and that of the affected bronchus was defined as the T-EtCO2. In other words, when the difference between EtCO2 of a specific bronchus and R-EtCO2 is more than T-EtCO2, this specific bronchus was diagnosed as the affected bronchus.

Statistical analysis

The detection powers of the EtCO2 test, balloon occlusion test and EtCO2 test plus balloon occlusion test were expressed as percentages. Different EtCO2 values between the affected bronchus and the non-affected bronchus, main bronchus and main carina were expressed as mean (95% confidence interval), and were assessed by independent sample Student’s t test. Statistical analysis was performed using the SPSS statistical software version 22.0 for Windows (SPSS, Inc., Chicago, IL, USA). A p value <0.05 was considered as statistically significant.

Results

Clinical characteristics of participants

A total of 28 male patients (64.7 ± 12.8 years old) with intractable pneumothorax were included in the present study. Of the 28 patients, 18 had right pneumothorax and 10 had left pneumothorax. Primary comorbidities of these patients were: chronic obstructive pulmonary disease (n = 18), old pulmonary tuberculosis (n = 4), pneumosilicosis (n = 5), lung cancer (n = 2), peripheral pleural fistula after left upper lobe wedge resection (n = 1) and idiopathic pneumothorax (n = 1).

Results of balloon occlusion and endo-bronchial EtCO2 test

The results for both the balloon occlusion test and the EtCO2 test are listed in Table 1. The balloon occlusion test successfully identified the affected bronchus in 18 cases and failed in 10 (positive rate 64.3%); the EtCO2 test showed reduction in EtCO2 values in 17 cases (positive rate 60.7%). The combined techniques failed in one case (positive rate 96.4%).

Table 1.

Power of endo-bronchial EtCO2 test and balloon detection.

EtCO2 test	Balloon occlusion test (n = 28)		Sum (%)[#]
(n = 28)	Positive (n)	Negative (n)
Positive	8	9	17 (60.7)*
Negative	10	1	11 (39.2)
Sum (%)	18 (64.3)*	10 (35.7)	28 (100)

There is a significant difference between positive rates of EtCO2 and balloon occlusion test (p = 0.041), tested by exact chi-square test. #Combination of the positive rate of the balloon occlusion test and the EtCO2 test was 96.4% (27/28).

Effectiveness of EtCO2 in patients with negative balloon occlusion results

Ten cases had negative results by the balloon occlusion test, nine of which showed positive results by the EtCO2 test; one case showed a negative result by both detection techniques.

Different values of EtCO2

The affected bronchus was successfully identified by EtCO2 test in 19 lobes or segments among 17 cases. The EtCO2 values recorded at different sites of the bronchial tree are listed and compared in Table 2. The EtCO2 values at the affected bronchus were significantly lower than those at the non-affected bronchus, main carina and main bronchus of the affected lung (all p < 0.05).

Table 2.

EtCO2 values among different sites.

Site of sampling	EtCO2 values (X¯ ± SD)	t-values (p) compared with affected bronchial	t-values (p) compared with non-affected bronchial
Affected bronchial	34.27 ± 8.49	–	−3.477 (p < 0.05)
Non-affected bronchial	41.45 ± 7.49	−3.337 (p < 0.05)	–
Main carina	40.00 ± 8.37	−2.102 (p = 0.042 < 0.05)	−0.652 (p = 0.517)
Main bronchus of affected lung	40.35 ± 8.24	−2.246 (p = 0.031 < 0.05)	−0.496 (p = 0.622)

Therapeutic results of bronchial occlusion

According to the results of the balloon occlusion test and the EtCO2 test, 27 patients underwent bronchial occlusion by the use of bronchial spigot, autologous blood and thrombin, or both. The overall successful rate was 24/28 (85.7%) (Figure 3). And air leak disappeared during the bronchoscopic procedure in all 24 successfully treated patients.

Figure 3.

Bronchial occlusion results: according to the combination detection results of the balloon occlusion test with EtCO2 detection, 27 patients underwent bronchial occlusion by the use of bronchial spigot, autologous blood or both. The successful rate of occlusion was 85.7%.

Discussion

Selective bronchial occlusion has been reported to be an effective technique for treating intractable pneumothorax.[1-4,6] Before bronchial occlusion can be performed, the affected bronchus must be located. Methods for identifying the affected bronchus include balloon detection, bronchography, Xe-133 detection and helium detection.[7] The balloon occlusion test is the most frequently used method, and it was associated with a positive detection rate of 85% in our previous study.[4] However, Sasada and colleagues reported that only 10 out of 23 patients showed apparent improvement in air leaks when treated with the balloon occlusion test.[8] The balloon occlusion test failed in >30% of cases of intractable pneumothorax in the present study. Reasons for failure of the balloon occlusion test might include multiple fistulas in different lobes and collateral ventilation between lobes,[9] neither of which can be detected when using a balloon catheter to occlude the orifice of a single lobar or segmental bronchus. Accordingly, there is a need for the development of a new technique to improve detection power in patients with multiple pleural fistulas in different lobes or collateral ventilation. In our study, the EtCO2 test was evaluated as a new technique for improving detection power, and we tried to establish a methodology for the test.

The endo-bronchial EtCO2 test was previous reported by Bhattacharyya and colleagues[10] as an effective technique for detecting the affected bronchus in patients with bronchopleural fistula. The authors stated that a particular affected bronchus was located when capnographic tracing showed rapid flattening of the curve to a straight line. In the present study, our tasks were to establish two standards: (1) the EtCO2 reference point; and (2) the threshold EtCO2 or T-EtCO2, which is the breakpoint of the diagnoses. First, the EtCO2 reference point: the EtCO2 sampled from this reference point (one or more of the following sites: main carina, main bronchus and other bronchus from the affected lung) is regarded as the standard or reference EtCO2 value (R-EtCO2) and can be compared with the EtCO2 value obtained from other points within the bronchial tree of the affected lung. Second, T-EtCO2 is the difference in EtCO2 between the affected bronchus and the reference point. In other words, T-EtCO2 allows us to determine the leading bronchus: when the difference in EtCO2 between a specific bronchus and the reference point is larger than the T-EtCO2, the specific bronchus can be labeled as the leading bronchus.

The EtCO2 values from the orifices of different bronchi may vary between patients because patients with intractable pneumothorax can have a variety of underlying diseases (e.g. chronic obstructive disease, old tuberculosis and pneumosilicosis) and have different gas exchange efficacies. This is why the establishment of R-EtCO2 is important. To date, no previous data on this topic have been reported. The main carina and main bronchus of the affected lung were candidate reference points in this study. There are significant differences in EtCO2 between the main carina and affected bronchus (p = 0.042), main bronchus and affected bronchus (p = 0.037), and non-affected bronchus (average EtCO2 of all bronchi other than the leading bronchus) and affected bronchus (p < 0.05). The EtCO2 values were similar between the main carina and main bronchus of the affected lung, main carina and non-leading bronchus, and main bronchus and non-leading bronchus. The above results suggested the main carina and main bronchus of the affected lung should be used as reference points.

Although there was a significant difference between the EtCO2 of the affected bronchus and the average EtCO2 of all non-leading bronchi, it is time-consuming to detect and calculate the EtCO2 values at the orifices of all the non-affected bronchi. Additionally, the EtCO2 from different non-affected bronchi may vary greatly. Thus, in clinical practice, it is not feasible to detect and calculate the average EtCO2 of the non-affected bronchus (R-EtCO2). The difference of EtCO2 between the main bronchus and the affected bronchus was 4.73 ± 2.10 (95% confidence interval: 3.80, 5.66) mmHg in the present study, suggesting that a specific bronchus can be determined to be the affected bronchus when its EtCO2 value is about 5 mmHg lower than that of the main bronchus of the affected lung. In other words, a 5 mmHg reduction in EtCO2 is considered to be the T-EtCO2 or diagnostic value for an affected bronchus according to our results. Our results demonstrated that the affected bronchus can be located in 60.7% of cases by the EtCO2 test and 64.3% of cases by the balloon occlusion test.

Theoretically, the balloon occlusion test is indicated for a single fistula without collateral ventilation, while the EtCO2 test is intended for multiple fistulas or fistulas with collateral ventilation. In the present study, the balloon occlusion test failed in 35.7% of cases of intractable pneumothorax. For this reason, the EtCO2 test was employed as a new technique for searching the affected bronchus. We found a positive detection rate of 60.7% with this new method, lower than that of balloon detection. However, as shown in Table 1, when we combined the EtCO2 test with balloon detection, the combined positive rate of the detection of the affected bronchus rose to 96.4% (27/28), indicating that these two techniques are complementary. The EtCO2 test is thus recommended to be an effective technique when the balloon occlusion test fails.

Some improvements in technical details are needed for the use of the trans-bronchial EtCO2 test in future clinical investigations. First, how to minimize the fluctuation of EtCO2 due to an unstable respiration state or cough; second, how to make the trans-bronchial EtCO2 test more feasible by shortening the procedure time; third, how to prevent the EtCO2 sampling catheter from becoming blocked with mucus. Further well-designed study is required to investigate these issues.

Some limitations of the present study should be mentioned. First, the sample size of the present study was small, with only 28 patients enrolled in the study. However, most patients with pneumothorax can be cured by intercostal drainage within 7 days – that is, the number of patients with intractable pneumothorax is relatively small. Multiple centers and a large sample size study is needed. Second, we did not conduct artery gas analysis for detecting PaCO2 when detecting EtCO2 from the affected bronchus. Third, the EtCO2 values were only detected in large bronchi rather than smaller bronchi, which may not be precise enough.

In summary, the EtCO2 test can be taken as a complementary technique in the event of failure of the balloon occlusion test. Combination of the EtCO2 test with the balloon occlusion test can significantly improve the positive detection rate in locating affected bronchi in patients with intractable pneumothorax. According to the result, the EtCO2 values measured at the main carina or main bronchus of the affected lung can be taken as the R-EtCO2 value for comparison with the EtCO2 value obtained at the orifice of the suspected leading bronchus. A particular bronchus can be determined to be the affected bronchus when its EtCO2 value is 5 mmHg less than the R-EtCO2 value.