Prediction of Ablation Volume in Percutaneous Lung Microwave Ablation: A Single Centre Retrospective Study

Round 1

Reviewer 1 Report

Paper is well written and study objectives are clear. However, I have following concerns which need to be addressed before acceptance of paper for publication.

1.) Table I: DAP entry for group 2 mentions mean value, which is outside of provided range of data.

2.) Line 233: Please correct “manual position” to “manual positioning” or “manual placement”.

3.) Line 240: On lines 100-102, authors state that software generates ablation volume based on MWA antenna manufacturer data. In that case, term “predictive ablation zone modelling” should be removed as word “modelling” suggests that actual computer model with Multiphysics was used to model and predict ablation zone. Instead, authors can use terms like “look-up table-based prediction of ablation zone”.

4.) Lines 255-256: Authors correctly point out, that actual ablation zone was observed to be smaller than what data from manufacturer handbook/experiments predicted. While this could have been caused by shrinkage as noted in paper, other factors come potentially in play as well. Examples are proximity of airway, vessel, and organ boundary to the ablation zone, which can deform ablation zone shape and deviate from lookup table of manufacturer. This point should be further elaborated on by mentioning that actual ablation zone can differ from VPS visualization due to the lookup table approach and that there is a need for development of fast personalized computational modelling, which would allow for real ablation zone estimation during procedure for further enhancement of procedure guidance. This should be cited, suitable references include modelling papers with patient specific geometries, such as recent modelling studies in liver, or sensitivity studies of MWA.

5.) Paragraph on lines 270-278: This paragraph is commenting on results in Table I. However, data in table I do not agree with arguments/findings in this paragraph.

Specific contradictions:

-    Group 1 has shorter fluoroscopy time and in following text, explanation is given that due to VPS, antenna is repositioned more times and patient rescanned à would not this lead to longer fluoroscopy time? Can authors comment on relation between shorter fluoroscopy time, which resulted in higher DAP? Also, would not higher number of repositions lead to longer procedure as opposed to data in Table I (slightly shorter procedure for group 1)?

-    Reposition times are roughly same for both groups (range 0-4, for 1 and 0-5 for group 2), which does not go well with provided note about more repositions for group 1 (it makes theoretically sense, but is not seen in data). Authors can explain this by saying, that repositioning is practically happening in group 2 as well as operator is trying to reach the centre of tumor with antenna or following some other guidelines for placement.

-     Text says that DAP is almost double than that of group 2. In Table 1, provided mean for DAP and group 2 is double than that of group 1 (relates to my first point).

Author Response

Dear Editors and Reviewers,

We thank you for considering our manuscript for publication and for addressing useful revision comments that would help us improve our work and reach the aim we set for it.

For clarity, we will answer each comment individually.

We willingly accepted and tried to fulfil every comment

1. Table 1: we apologize for the mistake in reporting the exact data of DAP and procedure time, the number belonged to a previous statistical analysis, we corrected the mistake both in the table and in the main text.
2. Line 233: we corrected the typo in “manual positioning”.
3. Line 240: we gratefully accepted the suggested expression “look-up table-based prediction of ablation zone”.
4. Lines 255-256: we appreciated the suggestion and further developed the point in the text, providing additional references.
5. Paragraph on lines 270-278

• Group 1: Total DAP is given by the sum of fluoroscopy DAP and CBCT DAP. Fluoroscopy is used for antenna positioning and CBCT is to evaluate the position of the antenna after each repositioning. CBCT gives a greater amount of radiation when compared to fluoroscopy. This explains why while in group 1 the DAP was almost double that in group 2, fluoroscopy time was significantly shorter. One can say that in group 2, the interventional radiologist that operates without the VPS takes more time to try to reach the tumour with the antenna, thus leading to a longer fluoroscopy time. Since in group 1 the VPS showed the target, fluoroscopy time is shorter, but more CBCTs are performed to assess exactly whether the tumour area is completely covered or not. This explains why in group 1 there is a higher DAP without a significantly greater number of repositioning. We further clarified this point also in the manuscript.
• We answered this point in 5.1.

As explained in the first point, we made a mistake in reporting data that were mixed up with results from previous statistical analysis, we apologize for that. 

We hope that our answers and corrections are satisfactory. We are extremely grateful for the opportunity to improve our manuscript and reach as many readers as possible.

Best regards,

The authors

Reviewer 2 Report

This paper describes Using cone-beam CT imaging  and volumetric prediction software (VPS) together to improve results with MVA of lung malignancies

It compares 10 pt with and 10 without VPS

This is a retrospective study. How were the patients - ie research group first or second 10 pts.

Figure one they describe the technique and say that the redline shows an insufficient expected ablation margin while green lines shown  complete coverage. I cannot see green. I only see red

Figure 2  in the text it says shows satisfactory positioning . That is not what figure 2 shows in the paper    Figure 2 in the paper is the workflow diagram

Are figure 3 and 4 using VPS

Figure 4 shows you what recurrence looks like.

but can you show us what an ablation looks like without VPS and see if you in retrospect can see the area that’s going to recur

I’m not sure how you can do statistical analysis on 20 patients

Conclusion is a little strong since there’s only 10 patients in each group.  One could say using software during MWA of lung malignancies may increase the efficiency of treatment but further studies have to be done

Author Response

Dear Editors and Reviewers,

We thank you for considering our manuscript for publication and for addressing useful revision comments that would help us improve our work and reach the aim we set for it.

For clarity, we will answer each comment individually.

We hope that our answers and corrections are satisfactory. We are extremely grateful for the opportunity to improve our manuscript and reach as many readers as possible.

Best regards,

The authors

Reviewer 2

We tried to fulfil every comment and answered those that we could not address in the current manuscript.

• We have been using the VPS for a year now, so we retrospectively selected 10 patients where the software was employed and 10 previous procedures where it wasn’t. We assured the two groups were as similar as possible in age, comorbidities, and tumour characteristics. We underlined that in the section “materials and methods”.
• We added arrows in figure 1 to make the green contour more easily recognized. We added a new figure to show how the red line – meaning incomplete ablation margin – looks like and pointed it out with a white arrow.
• Figure 2 (now Fig. 3 after we added a new figure) shows the procedural workflow, we corrected the reference in the text.
• In both figures 3 and 4 (now 4 and 5) VPS was used for the MWA; we added this information in the caption.
• Figure 4 (now 5) is the only example in group 1 where the residual disease was found on 1-month follow-up CECT. As these are only preliminary results, we decided to focus on the residual disease rather than recurrence. We will take your suggestion into account when more results will be out.
• As we pointed out in the limitations of the study our overall number of patients is low, so we were not able to obtain many statistically significant results. The aim of our study remains to show our preliminary results and encourage more future studies on the matter.

We underlined in our discussion that our results are preliminary and further studies need to be conducted.