Laser Interstitial Thermal Therapy in Patients with Newly Diagnosed Glioblastoma: A Systematic Review

Round 1

Reviewer 1 Report

Although LITT is becoming popular in treatment of unresectable GBM, evidence supporting its use are sparse. The systematic review by Viozzi et al is therefore well timed and relevant for the neuro-oncological field. The article is well constructed and the methodology used is appropriate. However some formal problems have to be corrected before its further consideration.

Line 22: OS and PFS, abbreviations not explained before in the text

Please mention in the introduction that LITT can be used together with standard radio-chemotherapy to give proper clinical context.

Line 63: which day of December 2020? Please specify

Line 77-80: not clear if the authors were provided with the proper information to extract the data from the published paper with pooled data, please clarify. In this regard please check line 121. 

Line 90: which authors? Please specify

Line 111: spelling error in “where reported in te articles”

Line 121: would have been possible to extract data just for the nGBM?

Figure 1: 6 = relevant data pooled- do you mean, from the text, “the 4 articles re-ported combined results for recurrent and primary glioblastoma and/or glioblastoma with 122 anaplastic astrocytoma” + 2 articles reporting on 126 cost-effectiveness not specific for nGBM)" ? Please double check on the numbers otherwise.

Line 135: please provide reference for the study prospectively designed.

Table 1. : Please explain what # and * refer to as you did in the following tables. I would recommend a reformatting of the table, it is quite difficult to read

Table 2. :

 title “GRADE assessment of the evidence” change with “GRADE assessment of quality of evidence”

supratentorial : mentioned only here, please remove or specify in the text if only supratentorial nGBM were included.

patient older than 18 years suffering from supratentorial: patient age is mentioned only here, please specify in the text if only adults patients were included.

The average point estimate: do you mean just “point estimate”? Also median is reported in the column

RCT: please define, abbreviation not mentioned before

The choice of the color is not good, it is difficult to read the text.

I suggest reformatting the table to ease the reader.

Line 167: please specify that age is reported in years

Line 186: which adjuvant treatment? Radiotherapy, temozolomide, lomustine, avastine? Not essential to define the exact number but give an idea.

Line 195, change 3 with three (beginning of a sentence)

Table 5: the title is misleading, not just complications are reported in the table, please change it.

Discussion: well organized and readable but I would like the author to mention also their clinical trial ongoing at the Radboud University that hopefully will give important information on the efficacy of LITT in nGBM.

Line 279-280: please remove instruction from MDPI

Author Response

Review 1:

Although LITT is becoming popular in treatment of unresectable GBM, evidence supporting its use are sparse. The systematic review by Viozzi et al is therefore well timed and relevant for the neuro-oncological field. The article is well constructed and the methodology used is appropriate. However some formal problems have to be corrected before its further consideration.

1. Line 22: OS and PFS, abbreviations not explained before in the text.

Thanks for noting this. We have revised this accordingly. Line 22-23: “Median overall survival (OS) ranged from 4.1 to 32 months and progression free survival (PFS) from 2 to 31 months.”

2. Please mention in the introduction that LITT can be used together with standard radio-chemotherapy to give proper clinical context.

As suggested, we have added a sentence to clarify this point (line 40-42): “Patients with unresectable glioblastoma could especially benefit from LITT, as LITT can be added to the standard adjuvant chemotherapy and radiotherapy to improve clinical outcomes”.

3. Line 63: which day of December 2020? Please specify.

As suggested, we have now specified the date of our last search (Line 67): “The last search was performed on 17 December 2020.” 

4. Line 77-80: not clear if the authors were provided with the proper information to extract the data from the published paper with pooled data, please clarify. In this regard please check line 121.

We believe that the unclarity about this point is generated from the term “included studies”: when we screened the articles, we found overlapping and we contacted authors (see answer to point 5 for specific names) for clarifications. If, based on authors’ answers, we found two or more articles showed patients overlap, we included the article with the most complete set of outcomes or with the largest population. To prevent confusion, we changed the word “included” in “screened” (line 81). Excluded articles are now also citated in the manuscript (line 83).

As some overlap could still be possible, we marked in the table’s studies with a # or a * when some overlapping could not be excluded.

5. Line 90: which authors? Please specify.

We contacted the following authors: Alireza M. Mohammadi, MD, Eric C. Leuthardt, MD, PhD, Albert H. Kim, MD, PhD, Ashish H. Shah, MD. Excluded articles are now also citated in the manuscript (line 83).

6. Line 111: spelling error in “where reported in te articles”.

Thanks for noting this. We changed this accordingly (line 117).

7. Line 121: would have been possible to extract data just for the nGBM?

We excluded studies where extraction of data for nGBM only was not possible. This sentence is now rephrased (Line 121-125): Based on title and abstract, 810 articles were excluded due to incompatibility with our eligibility criteria ( (…), studies regarding LITT in patients with a diagnosis other than nGBM (recurrent glioblastoma, metastases, radiation necrosis or other intracerebral tumours; studies regarding LITT in patients with epilepsy).

8. Figure 1: 6 = relevant data pooled- do you mean, from the text, “the 4 articles reported combined results for recurrent and primary glioblastoma and/or glioblastoma with 122 anaplastic astrocytoma” + 2 articles reporting on 126 cost-effectiveness not specific for nGBM)" ? Please double check on the numbers otherwise.

Figure 1 has been changed. To prevent any misunderstanding, we specified: 4= outcomes data pooled + 2=cost-effectiveness data not specific for nGBM (line 135).

9. Line 135: please provide reference for the study prospectively designed.

We now provide the references in the manuscript (reference 20, line 140)

10. Table 1. : Please explain what # and * refer to as you did in the following tables.

Changed accordingly: *: might include part of the same population. NR: not reported. NR^ data pooled from different populations. (line 169-170)

11. I would recommend a reformatting of the table, it is quite difficult to read Table 2.

Table 2 has been reformatted and we hope the reviewer agrees it is now easier to read. (Line 190-191)

12. Title “GRADE assessment of the evidence” change with “GRADE assessment of quality of evidence”.

The title has been changed as suggested by the reviewer (line 191).

13. Supratentorial : mentioned only here, please remove or specify in the text if only supratentorial nGBM were included.

We now specify in the text that we included studies on supratentorial glioblastoma (see line 74): “We included studies investigating LITT in supratentorial nGBM in adult patients (>18 years old)”

14. Patient older than 18 years suffering from supratentorial: patient age is mentioned only here, please specify in the text if only adults patients were included.

Please, see answer to point 13.

15. The average point estimate: do you mean just “point estimate”? Also median is reported in the column.

We indeed mean the point estimate here, and have changed it in the table accordingly (line 190-191).

16. RCT: please define, abbreviation not mentioned before The choice of the color is not good, it is difficult to read the text. I suggest reformatting the table to ease the reader.

We defined RCT accordingly (line 191). Please, see also answer to point 11. We changed the colour of the table with a more suitable one. (Line 190-191)

17. Line 167: please specify that age is reported in years.

As suggested, we now specify that age is reported in years, see line 201: “Mean age is reported in years”.

18. Line 186: which adjuvant treatment? Radiotherapy, temozolomide, lomustine, avastine? Not essential to define the exact number but give an idea.

Reporting of type of adjuvant treatment was very inconsistent. Where reported, most stated to have given standard chemotherapy and radiotherapy. Only one study (from Beaumont) reported in 2 patients lomustine and in 2 patients doxorubicine. We added: “The chosen adjuvant treatment was reported very inconsistently. Most of the studies report using standard chemotherapy with Temozolomide and radiation.” see line 221-223.

19. Line 195, change 3 with three (beginning of a sentence).

We changed the text following your suggestion, see line 232.

20. Table 5: the title is misleading, not just complications are reported in the table, please change it.

The title of table 5 has been changed in: “Table 5. Complications, tumour volumes, mortality and hospital stay” (line 249).

21. Discussion: well organized and readable but I would like the author to mention also their clinical trial ongoing at the Radboud University that hopefully will give important information on the efficacy of LITT in nGBM.

We have added (Line 329-334): “In line with these findings, we will perform a randomized pilot study to compare LITT to biopsy only in patients with unresectable glioblastoma, followed by adjuvant chemoradiotherapy according to the standard guidelines (Radboudumc, Nijmegen, Netherland [Clinicaltrials.gov ID: NCT04596930]. If data on safety and feasibility will be satisfying, we aim to perform a multicentre randomized controlled trial in the Netherlands to investigate cost-effectiveness of LITT.”

22. Line 279-280: please remove instruction from MDPI.

We changed the manuscript accordingly, see line 345-346.

Reviewer 2 Report

The authors present a comprehensive, systematic review or LITT for newly dx GBM. They clearly lay out their methodology and analysis and rationales. The results are limited by a lack of appropriate literature to include in their summary of current available data on the safety and effectiveness of LITT for newly dx GBM. They highlight the heterogeneity of the literature and that it can be difficult to draw true conclusions. Despite this, they have accomplished as best as possible, their goal of summarizing LITT for newly dx GBM. I feel that they had some opportunity to discuss LITT for newly dx GBM in comparison or relation to what is known for resected nGBM that they did not take advantage of. The real question that needs to be answered is "Does LITT provide similar benefits to OS and PFS as resection in newly diagnosed?". The tumor that will have LITT upfront at this time are usually lesions that would otherwise not have a resection (nGBM), inherently making analysis/comparison challenging. Therefore, it is necessary to look at historical reports for outcomes in nGBM with resection and adjuvant treatment. This is not ideal but helpful. The authors even state that unresectable tumors may benefit from LITT if they cannot go for resection. It is unlikely that a randomized controlled trial comparing LITT to open resection will occur for nGBM. The authors state that they do not recommend LITT for these patients outside of a trial which I disagree with.

There is certainly a paucity of data but I think the authors have an opportunity to explore the above a bit to add something to their discussion. A few other points:

-please reword first sentence, lines 35-36 for better clarity

-mentioning a comparison LITT to radiation is not relevant as these are not similar modalities intended to provide a similar therapy (such as surgery to remove it v. burning the tumor). It gives the reader a sense that the authors are not clear about treatments.

-they mention disruption of BBB but LITT disruption of this is not permanent

-the introduction can be reworked a bit to improve it, make it more directed and focused and to also remove lines 52-56 to methods

-line 111, should be "..in the articles."

-line 115, should add "...patients with newly diagnosed glioblastoma..."

-table 1 seems to have duplicate information and should be revised(bottom part is same as top)

-revise color scheme of table 2 to make more reader friendly  (text hard to see with dark color)

-also table 2, average point estimate for complications should be 33.7%

-table 3, Schroeder et al has an n of 3 but they do not list the distribution of these cases in the locations. Also, for Kamath et al, same table, they only account for 9 of 23 cases in location. If unknown, should state NA or NR.

-line 162, should be "Only two studies...."

-another interesting thing to report on, if data available would be laterality.

-table 4 and others, the time interval type is not specified (ie days, months, years). It is presumed to be months (table 4)but should be clarified in the table for the reader. Assume it is days for table 5...

-lines 226 and 242, should be unresectable from irresectable

-line 242, add "...glioblastoma as a treatment.."

-the paragraph from lines 242-247 is confusing for this reviewer. It seems as if multiple ideas are being thrown together that are not necessarily related. Example, once again, reference to radiation and how LITT could be performed multiple times unlike radiation. Also, the authors talk about not stopping chemotherapy which for newly diagnosed is irrelevant as we do not usually give neoadjuvant chemotherapy therefore, it is assumed they are talking about at recurrence. They also comment about how the majority of patients in their review could receive chemotherapy after LITT but I do not understand why- most patients would and it would be standard of care. They again mention  BBB permeability but it is transient (unknown for how long) and they should mention this thought with the reference that maybe there is theoretically enhanced tumor exposure to chemo with this if that is the point they are trying to make.

-in the next paragraph, starting with line 248 talking about complication rates, it is unclear if they are talking about LITT complication rates as a whole or specifically for tumors, just GBM, or just the nGBM cohort. Should be clarified and should compare similar (nGBM to nGBM) because factors such as tumor type, location, size, etc will have an impact on this.

Author Response

Review 2:

The authors present a comprehensive, systematic review or LITT for newly dx GBM. They clearly lay out their methodology and analysis and rationales. The results are limited by a lack of appropriate literature to include in their summary of current available data on the safety and effectiveness of LITT for newly dx GBM. They highlight the heterogeneity of the literature and that it can be difficult to draw true conclusions. Despite this, they have accomplished as best as possible, their goal of summarizing LITT for newly dx GBM. I feel that they had some opportunity to discuss LITT for newly dx GBM in comparison or relation to what is known for resected nGBM that they did not take advantage of. The real question that needs to be answered is "Does LITT provide similar benefits to OS and PFS as resection in newly diagnosed?". The tumor that will have LITT upfront at this time are usually lesions that would otherwise not have a resection (nGBM), inherently making analysis/comparison challenging. Therefore, it is necessary to look at historical reports for outcomes in nGBM with resection and adjuvant treatment. This is not ideal but helpful. The authors even state that unresectable tumors may benefit from LITT if they cannot go for resection. It is unlikely that a randomized controlled trial comparing LITT to open resection will occur for nGBM. The authors state that they do not recommend LITT for these patients outside of a trial which I disagree with.

There is certainly a paucity of data but I think the authors have an opportunity to explore the above a bit to add something to their discussion.

We strongly agree with the reviewer’s opinion about the necessity to answer the question whether LITT could provide similar results as standard of care. Unfortunately, given the lack of high-quality data, we were not able to answer this question. As evidence to perform LITT instead of surgical resection are lacking, LITT is for now especially used in patients with newly diagnosed glioblastoma when resection is not deemed feasible. Therefore, the most suitable population for outcomes comparison would be patients with a glioblastoma who undergo a biopsy only, followed by standard adjuvant chemotherapy and radiotherapy. In order to give our results more context, we have rewritten lines 273-277  in our discussion about the role of resection in survival: “As evidence to perform LITT instead of surgical resection in patients with newly diagnosed glioblastoma is lacking, LITT is for now used to treat patients where a resection is not deemed feasible. To compare these results to the current standard of care of patients with unresectable glioblastoma, we searched for literature investigating survival in patients undergoing a biopsy only, followed by adjuvant chemotherapy and radiotherapy (…).”

We started a randomized controlled pilot at our institution and hope to start a clinical trial soon to be able to answer this very interesting question (lines 334-339): “In line with these findings, we will perform a randomized pilot study to compare LITT to biopsy only in patients with unresectable glioblastoma, followed by adjuvant chemoradiotherapy according to the standard guidelines (Radboudumc, Nijmegen, Netherland [Clinicaltrials.gov ID: NCT04596930]. If data on safety and feasibility will be satisfying, we aim to perform a multicentre randomized controlled trial in the Netherlands to investigate cost-effectiveness of LITT.

A few other points:

1. Please reword first sentence, lines 35-36 for better clarity.

As suggested, we have rephrased the first sentences as follows (Line 36-42): ”Patients with a newly diagnosed glioblastoma who are not amenable for surgery and only receive chemoradiation miss the benefit of surgical resection. Survival is significantly worse for these patients as compared to those who undergo subtotal or gross total resection. Patients with unresectable glioblastoma could especially benefit from LITT, as LITT can be added to the standard adjuvant chemotherapy and radiotherapy to improve clinical outcomes.”

2. Mentioning a comparison LITT to radiation is not relevant as these are not similar modalities intended to provide a similar therapy (such as surgery to remove it v. burning the tumor). It gives the reader a sense that the authors are not clear about treatments.

We eliminated the comparison with radiotherapy, to prevent any misunderstanding (line 45).

3. They mention disruption of BBB but LITT disruption of this is not permanent.
   

We agree with the reviewer and have revised the text (line 47-49): “It has also been advocated that LITT could temporarily increase permeability for chemotherapeutic agents by disrupting the blood-brain-barrier”.

4. The introduction can be reworked a bit to improve it, make it more directed and focused and to also remove lines 52-56 to methods.
   

We have revised the introduction as suggested, see also answer to comment 1. We also removed line 55-59, as this is better described in the section methods (line 94-99).

5. Line 111, should be "..in the articles."

Thanks for noting this. We changed this accordingly (line 117).

6. Line 115, should add "...patients with newly diagnosed glioblastoma...".

We have added ”with newly diagnosed glioblastoma’’ as suggested  (Line 116).

7. Table 1 seems to have duplicate information and should be revised(bottom part is same as top).
   

Table 1 has been  revised accordingly (line 151-167).

8. Revise color scheme of table 2 to make more reader friendly (text hard to see with dark color)

Table 2 has been reformatted and we hope the reviewer agrees it is now easier to read. (Line 190-191)

9. Also table 2, average point estimate for complications should be 33.7% .

Thanks for noting this. We changed Table 2 accordingly (line 190).

10. Table 3, Schroeder et al has an n of 3 but they do not list the distribution of these cases in the locations. Also, for Kamath et al, same table, they only account for 9 of 23 cases in location. If unknown, should state NA or NR.

We added ‘(not further specified)’ and ‘lobar (14)’ where appropriate to table 3 to clarify this (line 200).

11. Line 162, should be "Only two studies...."
    

We corrected the spelling error (line 197).

12. Another interesting thing to report on, if data available would be laterality.

Following your suggestion, we added one column to table 3 about laterality (see line 200). Also, we added to the text: Laterality was reported in 5 studies for 28 patients (10 right side, 13 left side, 5 bilateral) (line 208-209).

13. Table 4 and others, the time interval type is not specified (ie days, months, years). It is presumed to be months (table 4)but should be clarified in the table for the reader. Assume it is days for table 5...

Indeed, survival is reported in months, age in years, hospital stay in months. We revised tables 4 and 5 accordingly. (Line 228 and 248)

14. Lines 226 and 242, should be unresectable from irresectable.

We have corrected this throughout the manuscript.

15. Line 242, add "...glioblastoma as a treatment.

This line has been rewritten, assuming it was meant line 243 (now line 275-278): “we searched for literature investigating survival in glioblastoma patients undergoing a biopsy only, followed by adjuvant chemotherapy and radiotherapy as a treatment”. 

16. The paragraph from lines 242-247 is confusing for this reviewer. It seems as if multiple ideas are being thrown together that are not necessarily related. Example, once again, reference to radiation and how LITT could be performed multiple times unlike radiation. Also, the authors talk about not stopping chemotherapy which for newly diagnosed is irrelevant as we do not usually give neoadjuvant chemotherapy therefore, it is assumed they are talking about at recurrence. They also comment about how the majority of patients in their review could receive chemotherapy after LITT but I do not understand why- most patients would and it would be standard of care. They again mention BBB permeability but it is transient (unknown for how long) and they should mention this thought with the reference that maybe there is theoretically enhanced tumor exposure to chemo with this if that is the point they are trying to make.

We have revised this paragraph (line 294-304): “LITT could offer selected patients with unresectable glioblastoma a treatment option, providing cytoreduction where surgery is impossible. Potential advantages of LITT comprise its minimal invasiveness, the possibility of treating deep lesions, the ability to repeat treatment and possibly chemotherapy could be started sooner after LITT then after surgery. The majority of patients included in our review could receive adjuvant chemotherapy and radiotherapy after LITT, which indicates they preserved a good KPS and that LITT did not interfere with the post-operative standard treatment. The role of LITT in transiently disrupting the permeability of the BBB and thus theoretically improving the response on chemotherapy has been suggested but not demonstrated yet (13).”

17. In the next paragraph, starting with line 248 talking about complication rates, it is unclear if they are talking about LITT complication rates as a whole or specifically for tumors, just GBM, or just the nGBM cohort. Should be clarified and should compare similar (nGBM to nGBM) because factors such as tumor type, location, size, etc will have an impact on this.

The complication percentage regards newly diagnosed GBM only. To make it clearer, we revised: “The mean complication rate of LITT in nGBM from all articles is 33.7%. In previously published studies, where results for LITT were pooled for different patient populations (newly and recurrent glioblastoma, metastases, epilepsy),  complication rates for LITT vary between 13 and 26%.”

It is hard to use this estimate for comparison, since factors such as tumour type, location, size, etc were poorly reported. We have made this more clear in line 309-315: “A possible explanation for this result is the focus of our systematic review on nGBM only.  Possibly the included studies reported more than average deep seated tumours (61.8% in thalamus, basal ganglia, corpus callosum, insula) explaining more neurological complications, but since data on tumour location, volume and complications were inconsistently reported, further analysis wasn’t possible and we can only speculate.”

Round 2

Reviewer 1 Report

the authors have provided a convincing revised version of the article.