TUS-EPIC: Thoracic Ultrasonography for Exclusion of Iatrogenic Pneumothorax in Post Transbronchial Lung Cryobiopsy—A Safe Alternative to Chest X-Ray

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The authors evaluated the utility of transthoracic ultrasonography (TUS) in excluding iatrogenic pneumothorax (PTX) after transbronchial lung cryobiopsy. Their results indicate that TUS is a rapid and reliable bedside modality for PTX detection, with the potential to decrease reliance on chest radiography and thereby minimize radiation exposure. The study is clinically relevant, as it suggests that routine use of TUS in this context may streamline post-procedural management, enhance patient safety, and optimize healthcare resources. Nevertheless, several points warrant further clarification.

1. Since ultrasonography is inherently operator-dependent and requires adequate expertise, could the authors elaborate on whether there is a learning curve for this technique?
2. In your study, the final diagnoses of the entire study population are not provided. We recommend that the authors present these details. Furthermore, it would be important to clarify whether there is any correlation between the occurrence of PTX and the underlying disease. To our knowledge, certain ILDs, such as idiopathic pulmonary fibrosis (IPF), are associated with a higher risk of PTX following TBLC. We suggest that the authors address this aspect in the discussion.
3. Among patients with concordant pre- and post-TBLC TUS findings, two eventually developed PTX. Could the authors clarify why PTX might have been difficult to detect in such cases? For example, could factors such as pleural adhesions, the specific biopsy site, or other patient-related conditions have contributed?
4. To our knowledge, delayed PTX can occur after TBLC. Regarding the 42 patients who were discharged without clinical suspicion of PTX, were any cases of PTX subsequently detected during outpatient follow-up?

Author Response

Comment 1: Since ultrasonography is inherently operator-dependent and requires adequate expertise, could the authors elaborate on whether there is a learning curve for this technique?

Response 1: We agree that this is an important point, and we have included this in our manuscript (Lines 280 - 286). As this study was done by a single operator, the internal generalizability may be limited, however there is internal consistency. In future, larger studies, analyzing operators with different degrees in thoracic ultrasound will help better define the learning curve for assessing post - processing IPTX.

 

Comment 2: In your study, the final diagnoses of the entire study population are not provided. We recommend that the authors present these details. Furthermore, it would be important to clarify whether there is any correlation between the occurrence of PTX and the underlying disease. To our knowledge, certain ILDs, such as idiopathic pulmonary fibrosis (IPF), are associated with a higher risk of PTX following TBLC. We suggest that the authors address this aspect in the discussion.

Response 2: Thank you for the comment. We have addressed this in the discussion of our manuscript (Lines 231 - 235). There is literature that demonstrates that those with underlying IPF are at an increased risk for biopsy related IPTX. Though it is hard to come up with definitive conclusions in this study, larger studies should further explore the relationship between ILD subtypes and IPTX risk.

 

Comment 3: Among patients with concordant pre- and post-TBLC TUS findings, two eventually developed PTX. Could the authors clarify why PTX might have been difficult to detect in such cases? For example, could factors such as pleural adhesions, the specific biopsy site, or other patient-related conditions have contributed?

Response 3:Thank you so much for the feedback. We have addressed this in the manuscript (Lines 262 - 265). There were 2 patients who developed  IPTX despite concordant pre - and post - TBLCB TUS findings, and it is likely that delayed IPTX is the cause. In our opinion, though, factors like pleural adhesions did not play a factor as TUS findings  of sliding lung would not have been present in the pre-TBLCB TUS evaluation.

 

Comment 4: To our knowledge, delayed PTX can occur after TBLC. Regarding the 42 patients who were discharged without clinical suspicion of PTX, were any cases of PTX subsequently detected during outpatient follow-up?

Response 4: Thank you for the comment. There were no cases of PTX that were subsequently detected during outpatient follow up. We have added a statement to clearly state this(Lines 209 - 210).

Reviewer 2 Report

Comments and Suggestions for Authors

1. At lines 60-61, authors say “This study evaluates TUS, for the first time, as an alternative to CXR for ruling out 60 IPTX related to TBLCB”: I think here they should better define this article’s research question: what do they want to explore? Usefulness of TUS in asymptomatic patients? Sensitivity for pnx diagnosis?

 

2. Author should clarify how long after the end of the procedure TUS was performed: only after 15 minutes or also later (e.g. in case of symptoms/ check independently on symptoms)? Was the operator who performed the biopsies the same who performed TUS?

 

3. Do the authors have data regarding symptoms in the 7 patients with discordant TUS findings (that were prescribed a chest XR)? Is it implicit that they were all asymptomatic? In that case, I think it should be made explicit. I think authors should explicit data regarding symptoms in patients that were prescribed a chest XR due to TUS findings. In particular, among the asymptomatic ones, how many pnx were found? Ultimately, what is the benefit of performing TUS in asymptomatic patients? I think this value would be clinically very useful and, thus, worth to be highlighted.

 

4. Regarding patients discharged after observation with negative TUS, I think authors should explicitly report on their final outcome. Was the patient who developed pnx after 12h the only one? Were there further pnx cases, maybe not requiring treatment?

 

5. As author report, NPV is high, but PPV stops at 28,6%, I think this should be reported in discussion/conclusion. This, also because on 4 pnx, 2 had symptoms (and thus would have undergone XR anyway, also in a setting that reserve imaging only to symptomatic patients). Provided that the 7 patients with discordant TUS were all asymptomatic, the interesting finding is that TUS is able to find pnx even in asymptomatic ones.

Author Response

Comment 1: At lines 60-61, authors say “This study evaluates TUS, for the first time, as an alternative to CXR for ruling out 60 IPTX related to TBLCB”: I think here they should better define this article’s research question: what do they want to explore? Usefulness of TUS in asymptomatic patients? Sensitivity for pnx diagnosis?​​

Response 1: Thank you for the comment. We clarified the sentence to better define the research question being asked (Line 60 - 63).

 

Comment 2:  Author should clarify how long after the end of the procedure TUS was performed: only after 15 minutes or also later (e.g. in case of symptoms/ check independently on symptoms)? Was the operator who performed the biopsies the same who performed TUS?

Response 2: Thank you for the comment. We have further clarified when TUS was performed (Line 108 - 111). It was performed within 15 minutes of extubation in all patients, and iIf patients were to develop symptoms suggestive of IPTX (chest pain, dyspnea, new hypoxia) during their 1.5 hours of  observation, a CXR was obtained to further assess.

 

Comment 3:  Do the authors have data regarding symptoms in the 7 patients with discordant TUS findings (that were prescribed a chest XR)? Is it implicit that they were all asymptomatic? In that case, I think it should be made explicit. I think authors should explicit data regarding symptoms in patients that were prescribed a chest XR due to TUS findings. In particular, among the asymptomatic ones, how many pnx were found? Ultimately, what is the benefit of performing TUS in asymptomatic patients? I think this value would be clinically very useful and, thus, worth to be highlighted. 

Response 3: We agree and have explicitly stated the patients with discordant TUS findings were asymptomatic and is seen in Figure 4. In regards to the data about symptoms in patients that were prescribed a chest XR due to TUS findings, including the symptomatic ones, we discuss how many IPTX were found. The answer is 2 and is found in figure 4 as well as the results section. Regarding the benefit of performing TUS in asymptomatic patients, we discuss this in the introduction.

 

Comment 4: Regarding patients discharged after observation with negative TUS, I think authors should explicitly report on their final outcome. Was the patient who developed pnx after 12h the only one? Were there further pnx cases, maybe not requiring treatment?

Response 4: Thank you for the comment. We have clarified in the manuscript. (Line 179 - 183) Only 1 patient developed IPTX, 12 hours later. No additional chest imaging was obtained in the remaining patients as none manifested symptoms suggestive of a clinically - relevant delayed IPTX upon evaluation in the ambulatory interventional pulmonology clinic.

 

Comment 5: As author report, NPV is high, but PPV stops at 28,6%, I think this should be reported in discussion/conclusion. This, also because on 4 pnx, 2 had symptoms (and thus would have undergone XR anyway, also in a setting that reserve imaging only to symptomatic patients). Provided that the 7 patients with discordant TUS were all asymptomatic, the interesting finding is that TUS is able to find pnx even in asymptomatic ones.

Response 5: We agree. We have  elaborated on this in our discussion and will highlight the PPV. Thank you.

Reviewer 3 Report

Comments and Suggestions for Authors

This manuscript addresses an important and clinically relevant question: whether thoracic ultrasound (TUS) can safely replace routine chest X-ray (CXR) for excluding iatrogenic pneumothorax (IPTX) after transbronchial lung cryobiopsy (TBLCB). The topic is timely, given the increasing use of TBLCB and the need to optimize post-procedural safety protocols while minimizing radiation exposure. The study is well-written overall, with a clear methodology and promising findings. However, several issues require clarification and refinement before the manuscript can be considered for publication.

Major Concerns

1. Study design limitations
• Single-center, single-operator design restricts generalizability.
• Exclusion of patients with nonspecific pre-procedural TUS findings may introduce selection bias and artificially increase the NPV. These limitations should be emphasized more clearly in both the Discussion and Conclusion.
2. Diagnostic accuracy reporting
• The manuscript reports only the NPV. Sensitivity, specificity, and positive predictive value should also be calculated and reported, as these would provide a more complete assessment of TUS performance.
3. Figures and tables
• Figure 4 is difficult to interpret and should be simplified or redesigned for clarity.
• Table 2 would benefit from clearer labeling (e.g., “Concordant TUS” vs. “Discordant TUS” rather than TUS (+)/TUS (−)).
4. Pneumothorax timing
• The occurrence of a delayed symptomatic IPTX is an important finding that deserves stronger emphasis. It highlights the need for clinical vigilance and careful patient instructions even when TUS findings are negative.

Recommendation

This is a valuable contribution to the field, but revisions are required before publication. I recommend major revision with particular attention to:

• Reporting full diagnostic accuracy metrics,
• Expanding the limitations section,
• Clarifying the impact of delayed IPTX, and
• Improving figure/table clarity.

best regards

Author Response

Comment 1: Single-center, single-operator design restricts generalizability. Exclusion of patients with nonspecific pre-procedural TUS findings may introduce selection bias and artificially increase the NPV. These limitations should be emphasized more clearly in both the Discussion and Conclusion.

Response 1: Thank you for the comment. The patients with non-specific pre-procedural TUS findings were excluded because these findings rendered the postprocedural TUS to be unreliable, as the operator would be unable to determine if the findings after biopsy truly represent a pneumothorax vs. false positive conditions. We did clarify how selection bias may have artificially increased the NPV in the discussion (Line 286 - 291).

 

Comment 2: The manuscript reports only the NPV. Sensitivity, specificity, and positive predictive value should also be calculated and reported, as these would provide a more complete assessment of TUS performance.

Response 2: Thank you for the comment. We have added the calculations in the results section (Lines 190 - 192).

 

Comment 3: Figure 4 is difficult to interpret and should be simplified or redesigned for clarity. Table 2 would benefit from clearer labeling (e.g., “Concordant TUS” vs. “Discordant TUS” rather than TUS (+)/TUS (−)).

Response 3: We have further simplified Figure 4 and adjusted the labeling for Table 2.

 

Comment 4: The occurrence of a delayed symptomatic IPTX is an important finding that deserves stronger emphasis. It highlights the need for clinical vigilance and careful patient instructions even when TUS findings are negative.

Response 4: Thank you for the comment. We discuss this in the discussion and conclusion of the manuscript.

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

Authors sufficiently clarified and edited the previously raised points.

Author Response

Reviewer 2: Authors sufficiently clarified and edited the previously raised points.

- Thank you so much!

Reviewer 3 Report

Comments and Suggestions for Authors

Dear authors, 

The article is clearly written, the methodology is adequately described, and the findings are relevant to clinical practice. However, some methodological limitations and interpretative issues need to be addressed to strengthen the manuscript and support its conclusions.

Some sugestions:

Table 2 and Figure 4 could be improved for clarity. Consider adding additional metrics (e.g., time to IPTX, stratification by underlying disease).

Including ROC curves or additional diagnostic performance measures would enhance the robustness of the results.

Expand the discussion on existing evidence supporting TUS over CXR and position your findings within this broader context.

 

best regards

Author Response

The article is clearly written, the methodology is adequately described, and the findings are relevant to clinical practice. However, some methodological limitations and interpretative issues need to be addressed to strengthen the manuscript and support its conclusions.

Some sugestions:

Table 2 and Figure 4 could be improved for clarity. Consider adding additional metrics (e.g., time to IPTX, stratification by underlying disease). Including ROC curves or additional diagnostic performance measures would enhance the robustness of the results. Expand the discussion on existing evidence supporting TUS over CXR and position your findings within this broader context.

- Thank you so much for your feedback. To further enhance the robustness of our results, we included the area under the curve. We are hoping this helps strengthen the utility of ultrasound in our study. We also have expanded our discussion with existing evidence to support the use of TUS over CXR. In regards to clarifying table 2 and figure 4, the time to IPTX is mentioned in figure 4 as part of our protocolized study.