State of the Art in Combination Immuno/Radiotherapy for Brain Metastases: Systematic Review and Meta-Analysis

Round 1

Reviewer 1 Report

The present work is a metanalysis about the role and impact of the immunotherapy treatment combined with radiotherapy in the brain metastasis management. This is a very interesting topic and cover an unmet clinical need. The methods employed sound robust and solid, the results are well structured and presented, and personally, I have enjoyed reviewing this manuscript. However, I have several comments and questions, most of them, minors.

Introduction

1. Line 49: I don’t understand the real meaning of the first sentence when authors use the adverb ‘often’. In my understanding, the development of brain mets always implies progression.

Methods

1. The authors mention the use of Cochrane Collaboration tool to assess the quality of the studies, but in the selected studies to be analyzed is only reported the NOS. In my opinion, the assessment of these quality measures are not well explained in meths or it’s missed one of them.
2. In the same line, the agreement between reviewers is only assessed in the data extraction or also in the evaluation of the study quality properties? This is also an important point to clarify.

Results

1. In the figure 1 is missing the reason to eliminate 5 studies in the first (identification) step. According the final step of the figure, the authors identify 32 studies, but in the text they mention 28 (line 146). Please, clarify the discrepancy.
2. Table 1. I don’t understand the label of the column 7 (Brain mets (%)). What means this %?, and what means size? The maximum diameter of the mets to be included? the mean of the diameters? Please, clarify in footnotes.
3. Table 1. Which criteria of BM response were used in the studies? This point should have to be mentioned.
4. Line 177: it’s also no clear the way how is expressed the time between RT and IT. What means the range in %?, and all including the 100%?
5. Paragraph 187-197: The toxicities that implies dose reductions or withdrawing treatments have to be categorized as severe toxicities, therefore toxicities grade 3 has to be considered as severe. It would be interesting for the reader know not only the percentage range of the toxicities reported in the studies, also the weighted arithmetic mean. Regarding the radionecrosis, also it should be interesting know the time when this adverse event appears, and add some discussion/comments/clarifications about the possibility if the events recorded as necrosis can be pseudoprogresions or how discriminate the studies the progression from the pseudos if the patients were taken IT.
6. In the survival analysis, could it be possible do a survival subanalysis only including the highest quality studies, and by the 2 main histological subtypes (melanoma and NSCLC), and by the radiotherapy technique employed (WBRT and SRS)? The first, just to ensure that the present results go in the same direction and the second to explore if the effect on survival is class tumor dependent. The third should be also interesting because melanoma is a known radioresistant tumor.
7. Subheading 6. Lines 214-221. How many studies were included in this subanalysis?. This paragraph is difficult to follow and understand, needs improvement.

Discussion

1. Line 238: typo error.
2. Several metanalysis has been performed in the past years about the same topic. It should be interesting extend the discussion about what add to the medical knowledge the present work with regard to the previous studies (other than the number of studies and total patients included)
3. In the limitations or in the discussion, I think that it has to be reinforced the message that the studies can be biased by the unknown mutation status of the melanomas and NSCLC tumors. On the other hand, if in the limitations the authors recognize that the results cannot be generalized due to the overrepresentation of melanomas in the selected studies, why in the conclusion is included the validity of this results for lung cancer?
4. This work and the previous metanalysis pointed out to a benefit of the combination of RT and IT with regard RT. However, we don’t know if RT+IT is superior or equivalent to IT alone, especially taken into account the potential radiioresistence of melanomas, and it’s excellent response to the IT. I think that this point should have to be also discussed.

Author Response

Dear Reviewer

Thank you for very useful comments. Please find the point-by-point answers. Correction are made in the revised manuscript.

Sincerely Yours

Salvator Cozzi

Introduction

1. Line 49: I don’t understand the real meaning of the first sentence when authors use the adverb ‘often’. In my understanding, the development of brain mets always implies progression.

We meant it usually accompanies “extra”cranial progression of the disease. Changes are made in the manuscript.

Methods

1. The authors mention the use of Cochrane Collaboration tool to assess the quality of the studies, but in the selected studies to be analyzed is only reported the NOS. In my opinion, the assessment of these quality measures are not well explained in meths or it’s missed one of them.

We aimed to mention the quality assessment tools that we use generally for meta-analysis. For observational studies the most commonly used tool in the literature is Newcastle-Ottawa Scale and we used this tool. However we mentioned the Cochrane Collaboration’s tool as a general knowledge. But NOS was more suitable for our data (table ). Because of this confusion, I removed the first one.

1. In the same line, the agreement between reviewers is only assessed in the data extraction or also in the evaluation of the study quality properties? This is also an important point to clarify.

This is a very important point. Any disagreement between reviewers was resolved by discussion in whole study team. I included this point into the manuscript.

Results

1. In the figure 1 is missing the reason to eliminate 5 studies in the first (identification) step. According the final step of the figure, the authors identify 32 studies, but in the text they mention 28 (line 146). Please, clarify the discrepancy.

It is written in the table that 5 studies were copies. I added another box to the right and underscored that these 5 studies were duplicates.

Of final 32 studies, 4 were excluded because of imperfection in outcome reports. That’s why “28 studies” is mentioned in the text. The figure is edited to eliminate this confusion

1. Table 1. I don’t understand the label of the column 7 (Brain mets (%)). What means this %?, and what means size? The maximum diameter of the mets to be included? the mean of the diameters? Please, clarify in footnotes.

Column 7 shows the average number of metastases. In some studies percent of patients who had 1, 2, or more metastases are reported. These data are summarized in the table.

About the column of “Size”, I admit that some information has been missed. Because these numbers may differ in meaning. I Edited the table to make them clarified.

1. Table 1. Which criteria of BM response were used in the studies? This point should have to be mentioned.

Almost all studies had used response evaluation criteria in solid tumors (RECIST) and some studies used also immune-related response criteria (irRC). I added this to the table 1.

1. Line 177: it’s also no clear the way how is expressed the time between RT and IT. What means the range in %?, and all including the 100%?

It tried to show the range of diversity between different studies. That was unnecessary as they are already illustrated in table 2. So that was removed from the document.

1. Paragraph 187-197: The toxicities that implies dose reductions or withdrawing treatments have to be categorized as severe toxicities, therefore toxicities grade 3 has to be considered as severe. It would be interesting for the reader know not only the percentage range of the toxicities reported in the studies, also the weighted arithmetic mean. Regarding the radionecrosis, also it should be interesting know the time when this adverse event appears, and add some discussion/comments/clarifications about the possibility if the events recorded as necrosis can be pseudoprogresions or how discriminate the studies the progression from the pseudos if the patients were taken IT.

It is correct. G3 toxicity should be considered as severe. So, I corrected it in the text. All information about toxicities that could be accessed through full text articles are presented in table 2.  A more detailed discussion about radionecrosis is added into the section regarding toxicities.

1. In the survival analysis, could it be possible do a survival subanalysis only including the highest quality studies, and by the 2 main histological subtypes (melanoma and NSCLC), and by the radiotherapy technique employed (WBRT and SRS)? The first, just to ensure that the present results go in the same direction and the second to explore if the effect on survival is class tumor dependent. The third should be also interesting because melanoma is a known radioresistant tumor.

It is a very good idea. Unfortunately, in our study it was not possible due to small number of high-quality data in each subgroup. I will consider it in future works.

1. Subheading 6. Lines 214-221. How many studies were included in this subanalysis? This paragraph is difficult to follow and understand, needs improvement.

Thank you for this comment. Revision was made in the manuscript.

Discussion

1. Line 238: typo error.
2. Several metanalysis has been performed in the past years about the same topic. It should be interesting extend the discussion about what add to the medical knowledge the present work with regard to the previous studies (other than the number of studies and total patients included)

Previous meta-analyses were smaller with fewer patients analyzed. For example, in Petrelli et al. 754 patients from 13 studies were finally selected for statistical analysis (fig 2 in their paper). Effect size is not analyzed in most of them. I also added Rulli et al. study (cancer 2019), they had included target therapy as well as immunotherapy. I edited the text so that all similar studies come together and be discussed in more detail.

1. In the limitations or in the discussion, I think that it has to be reinforced the message that the studies can be biased by the unknown mutation status of the melanomas and NSCLC tumors. On the other hand, if in the limitations the authors recognize that the results cannot be generalized due to the overrepresentation of melanomas in the selected studies, why in the conclusion is included the validity of this results for lung cancer?

Thank to your suggestion, I added this point to limitations. Moreover, I revised the conclusion accordingly.

1. This work and the previous metanalysis pointed out to a benefit of the combination of RT and IT with regard RT. However, we don’t know if RT+IT is superior or equivalent to IT alone, especially taken into account the potential radiioresistence of melanomas, and it’s excellent response to the IT. I think that this point should have to be also discussed.

This is a good question, but it is difficult to answer with the current available studies, because radiotherapy is the standard of treatment in brain metastases and it is not easy to find enough studies, that have our eligibility criteria, in which patients with brain metastasis are treated with immunotherapy alone. But it is a good subject to work on and design future studies. I added this comment to our conclusion.

Author Response File: Author Response.docx

Reviewer 2 Report

Dr Najafi et al. performed a systematic review of the literature focusing on studies evaluating the combination of radiotherapy (either WBRT or SRS) with immunotherapy mainly in patients with advanced melanoma. The criteria used for selecting the studies were well defined and they ended with 28 studies (1,465 patients) that were included in the metanalysis. They focused the analysis on the treatment toxicity and overall survival. They claimed that combination of immunotherapy plus radiotherapy was superior to radiotherapy alone. However, all the studies included not only lack a randomized design but also are retrospective. I am concerned about the strong bias behind those studies that may have significant impact on the external and internal validity of this study. I consider that the conclusions should be tone down since the study design does not allow to draw any robust conclusion about the superiority of brain radiotherapy combined with immunotherapy over radiotherapy alone. Indeed, the most relevant question would be to know whether brain radiotherapy plus immunotherapy is superior to immunotherapy alone. In addition, the assessment of long-term toxicities, especially neurocognitive impairment, has been neglected in many studies assessing brain radiotherapy alone or combined with systemic therapies.

Table 2 should be more informative and provide per each study the frequency of G3-4 toxicity. The rate of radionecrosis is considerably high in some studies combining radiotherapy and immunotherapy and this was not discussed by the authors.

Author Response

Dear Reviewer

Thank you for very useful comments. Please find the point-by-point answers. Correction are made in the revised manuscript.

Sincerely Yours

Salvator Cozzi

Dr Najafi et al. performed a systematic review of the literature focusing on studies evaluating the combination of radiotherapy (either WBRT or SRS) with immunotherapy mainly in patients with advanced melanoma. The criteria used for selecting the studies were well defined and they ended with 28 studies (1,465 patients) that were included in the metanalysis. They focused the analysis on the treatment toxicity and overall survival. They claimed that combination of immunotherapy plus radiotherapy was superior to radiotherapy alone. However, all the studies included not only lack a randomized design but also are retrospective. I am concerned about the strong bias behind those studies that may have significant impact on the external and internal validity of this study. I consider that the conclusions should be tone down since the study design does not allow to draw any robust conclusion about the superiority of brain radiotherapy combined with immunotherapy over radiotherapy alone.

I totally agree with this comment and therefore, I have mentioned tis issue in the limitation part. However, retrospective studies were the only available studies for us. We did not insist on the conclusion and mentioned the need for studies with a more powerful design and included this in our conclusion. I have added this statement in the conclusion section for emphasis.

Indeed, the most relevant question would be to know whether brain radiotherapy plus immunotherapy is superior to immunotherapy alone.

This is a good question, but it is difficult to answer with the current available studies, because radiotherapy is the standard of treatment in brain metastases and it is not easy to find enough studies, that have our eligibility criteria, in which patients with brain metastasis are treated with immunotherapy alone. But it is a good subject to work on and design future studies. I added this comment to our conclusion.

 In addition, the assessment of long-term toxicities, especially neurocognitive impairment, has been neglected in many studies assessing brain radiotherapy alone or combined with systemic therapies. Table 2 should be more informative and provide per each study the frequency of G3-4 toxicity.

This is a good point about our study. Neurocognitive effects of radiotherapy are well reported in many studies, however the effect of combined immunotherapy, which is the focus here, on cognition is lacking in available studies. So, we could not include it in our analysis. I added this shortage in our limitation section. For studies that had provided such toxicity rates, the information is included in the table 2 in the last column. e.g., Hubbeling et al., Schapira et al., Nardin et al, etc. Unfortunately, some studies did not mention toxicity rates. So, we have presented all we could get access to.

The rate of radionecrosis is considerably high in some studies combining radiotherapy and immunotherapy and this was not discussed by the authors.

This is a very important point that was missed in our discussion. I included a paragraph in the discussion addressing this issue.

Author Response File: Author Response.docx

Reviewer 3 Report

Najafi et al present a review/meta-analysis of 32 (or is it 28…?) retrospective papers on the topic of combine immune/radiotherapy for brain metastases. They present the rationale, the method and analyses of survival and toxicity. They conclude that combined IR has prolonged OS and improved mortality rates with acceptable toxicity.

In general, during the text, statements regarding BM and cancer in general are mixed and un-clear. “The advent of new immunotherapeutic treatments has made a revolution in the treatment of these patients…” – meaning BM patients or treatment of the primary cancers? The effect of immunotherapy in BM is unclear – otherwise this review would be obsolete? So I guess the statement is meant for treatment of primary cancers – this should be written more clearly throughout the manuscript.

The manuscript should be thoroughly revised for grammar, and improved and clear language. Example: Line 241-242, this is poorly understandable. Please clearify. Line 333-334: “…patient characteristics were different from DS-GPA and Biology.” What does this mean?

SBRT/SRS/SRT is used as abbreviation for stereotactic intracranial radiotherapy. Please use a consistent form of abbreviation. SBRT is used as abbreviation for stereotactic radiotherapy. SBRT is mostly used for Stereotactic Body radiotherapy. I suggest using either SRT or SRS (for stereotactic radiosurgery) – SRT most often used when more than one fraction of stereotactic radiotherapy is used, SRS when one fraction is used.

The Methods section is well written in my opinion, but I am not an expert in meta-analysis as a method.

Abstract:

Line 39: “The addition of IT to RT compared to RT alone has a hazard ratio between 34% and 44%.” HR for what? Please specify.

Page 1 line 36: “…the addition of IT to RT was significantly better than RT alone..”. This is very unspecific and should be omitted.

Reference 4 in the introduction – I dont understand the reference in relation to BM treatment – is it a general comment on treatment for oligometastatic disease or is it meant to relate to BM treatment? If the latter, it is misplaced. If the former, it should be specified.

Introduction:

The introduction can be trimmed and made more consistent.

The studies from Goldberg (ref 15) and Long (rf 16) are on IT alone, not in combination with RT. The statement in line 79 is inconsistent with these references. Please clearify.

Results:

Is the final analysis made from 32 or 28 papers? The figure 1 and text are not consistent. Is there only one relevant study published in 2019 and none after that?  Why not cite Rauschenberg (Eur J cancer, 2019)? Or Lannier Neurooncol Pract 2019)?  Minntti (J IUmmunother Cancer 2019)? Are there others from 2019 until 2021?

Line 201 – “Stereotactic Body radiation therapy…treatment of choice in BMs”. I believe it is wrong to write stereotactic body RT when it comes to brain RT. SBRT is RT outside the brain – intracranial stereotactic RT is either named SRS or SRT (alt. fractionated SRT, fSRT). Please clearify/correct.

Line 203 – “SRS versus SBRT” or should it be “SRS vs. WBRT”…?

Discussion:

In terms of timing, RT seems to have the best effect on the result when performed before or simultaneously with immunotherapy. What is the definition of “before” , “simultaneously”/concurrent” and “after”. Is “before” any time before or immediately (and what is that..) before RT? Patients developing BM during IT treatment must be regarded as belonging to a totally different prognostic group than “IT naïve” patients diagnosed with BM.

The authors should discuss the time-bias phenomenon of IT treatment after BM – do patients live longer because they receive  IT treatment or do patients receive IT because they live long enough to receive it…?

Also, the authors should mention/discuss whether the improved OS in combining IT and RT results from better intracranial effect or effect on extracranial disease, or both. They should mention that extracranial disease status may have more impact on survival after BM than perhaps previously emphasized.

The meta-analyses from Petrelli and Lehrer are cited – these meta-analyses should be discussed together, not in separate paragraphs.  The authors should comment on what their current review/meta-analysis add to these three reviews, given that no studies but 1 is included from 2019 and to date in the analyses in this manuscript? Why is Rulli et al (Cancer 2019) not cited?  

An increased risk of radionecrosis (RN) is one important side-effect that should be considered when combining intracranial RT and IT. I think this should be discussed further in the Discussion, and also considering the fact that RN may occur increasingly after 1-2 years and later, this should be discussed according to the follow-up times in papers cited in the Toxicities section.

All included papers were retrospective. There are no published prospective or randomized studies on IT and BM RT, as far as I know. The authors should discuss further how this limits the results in their analyses, they only mention this in a summerical manner. What studies, in the opinion of the authors, should be conducted to further enlighten us on this topic, and they should provide an overview of examples of clinical trials that are underway/being conducted (from clinicaltrials.gov or other sources).  

The authors should be more clear and specific as to what their meta-analaysis adds to those already published (see above). Also - What is the recommendation from the authors after performing this meta-analysis – the “take home message”? Should we strive to combine IT and RT for BM? And how? Or is the recommendation simply that further studies are needed? The latter we know – we do need further studies. But can this current analysis guide us while awaiting the results if such studies?  

Author Response

Dear Reviewer

Thank you for very useful comments. Please find the point-by-point answers. Correction are made in the revised manuscript.

Sincerely Yours

Salvator Cozzi

Najafi et al present a review/meta-analysis of 32 (or is it 28…?) retrospective papers on the topic of combine immune/radiotherapy for brain metastases. They present the rationale, the method and analyses of survival and toxicity. They conclude that combined IR has prolonged OS and improved mortality rates with acceptable toxicity.

For meta-analysis, only 28 studies could be included. I have edited the figure 1 so that in become clear why 4 studies were excluded.

In general, during the text, statements regarding BM and cancer in general are mixed and un-clear. “The advent of new immunotherapeutic treatments has made a revolution in the treatment of these patients…” – meaning BM patients or treatment of the primary cancers? The effect of immunotherapy in BM is unclear – otherwise this review would be obsolete? So I guess the statement is meant for treatment of primary cancers – this should be written more clearly throughout the manuscript.

It is meant for treatment of primary cancers. I admit that there is some ambiguity in the text, so, we revised the document accordingly.

The manuscript should be thoroughly revised for grammar, and improved and clear language. Example: Line 241-242, this is poorly understandable. Please clearify. Line 333-334: “…patient characteristics were different from DS-GPA and Biology.” What does this mean?

I totally agree that, there were weaknesses in language of the manuscript. We revised the whole manuscript accordingly. DS-GPA (diagnosis-specific graded prognostic assessment) score is a grading system developed for stratification of prognosis in brain metastases. I modified the text to become clearer.

SBRT/SRS/SRT is used as abbreviation for stereotactic intracranial radiotherapy. Please use a consistent form of abbreviation. SBRT is used as abbreviation for stereotactic radiotherapy. SBRT is mostly used for Stereotactic Body radiotherapy. I suggest using either SRT or SRS (for stereotactic radiosurgery) – SRT most often used when more than one fraction of stereotactic radiotherapy is used, SRS when one fraction is used.

Thank you for this comment. I corrected the manuscript accordingly.

Abstract:

Line 39: “The addition of IT to RT compared to RT alone has a hazard ratio between 34% and 44%.” HR for what? Please specify.

That is a good point. Hazard ratio for addition of immunotherapy to radiotherapy was 39%. 34% to 44% is the 0.95 confidence interval for this hazard ratio. I made corrections in document to become clearer.

Page 1 line 36: “…the addition of IT to RT was significantly better than RT alone..”. This is very unspecific and should be omitted.

Thank you for this comment, I revised the manuscript accordingly.

Reference 4 in the introduction – I dont understand the reference in relation to BM treatment – is it a general comment on treatment for oligometastatic disease or is it meant to relate to BM treatment? If the latter, it is misplaced. If the former, it should be specified.

This part of instruction is trying to mention the paradigm shift that occurred toward local therapy in brain metastatic settings with one to few metastases. Early studies that showed a benefit from metastatectomy and later studies that generalized this concept toward radiosurgery in oligometastases. I changed the sentences’ structure a little bit so that it can be more understandable.

Introduction:

The introduction can be trimmed and made more consistent.

We made an overall language revision on the manuscript.

The studies from Goldberg (ref 15) and Long (rf 16) are on IT alone, not in combination with RT. The statement in line 79 is inconsistent with these references. Please clearify.

I admit that this part of introduction could be written in a more fluent way. We started with RT and then explained studied that exploited IT for treatment of metastasis and finally discussed their combination. The sequence was not clear, so we have made pertinent modification.

Results:

Is the final analysis made from 32 or 28 papers? The figure 1 and text are not consistent. Is there only one relevant study published in 2019 and none after that?  Why not cite Rauschenberg (Eur J cancer, 2019)? Or Lannier Neurooncol Pract 2019)?  Minntti (J IUmmunother Cancer 2019)? Are there others from 2019 until 2021?

32 studies passed the different stages of screening that are presented in the table. As it is mentioned in the txt, metanalysis was performed on 28 studies that fulfilled the requirement for statistical analysis (figure 2). I revised the figure and mentioned why 4 studies were excluded in the last stage. We have searched the literature according to the mentioned keywords. Some studies might have been omitted because either not being in the search results with the selected search terms or not fulfilling our secondary selection criteria.

Line 201 – “Stereotactic Body radiation therapy…treatment of choice in BMs”. I believe it is wrong to write stereotactic body RT when it comes to brain RT. SBRT is RT outside the brain – intracranial stereotactic RT is either named SRS or SRT (alt. fractionated SRT, fSRT). Please clearify/correct.

Line 203 – “SRS versus SBRT” or should it be “SRS vs. WBRT”…?

Thank you for these comments. We have corrected the text according to your comments.

Discussion:

In terms of timing, RT seems to have the best effect on the result when performed before or simultaneously with immunotherapy. What is the definition of “before” , “simultaneously”/concurrent” and “after”. Is “before” any time before or immediately (and what is that..) before RT? Patients developing BM during IT treatment must be regarded as belonging to a totally different prognostic group than “IT naïve” patients diagnosed with BM.

The authors should discuss the time-bias phenomenon of IT treatment after BM – do patients live longer because they receive  IT treatment or do patients receive IT because they live long enough to receive it…?

Different authors have different definitions for those. Chen et al defines concurrent as “within two weeks before or after SRS/SRT” non-concurrent meant more than 2 weeks apart. Most studies consider 30 days as being concurrent or not. This definition is integrated into the manuscript.

There are some limitations regarding definition of concurrent and non-concurrent and also time-bias. These limitations could not be avoided and therefore we added this issue into limitation section of the manuscript.

Also, the authors should mention/discuss whether the improved OS in combining IT and RT results from better intracranial effect or effect on extracranial disease, or both. They should mention that extracranial disease status may have more impact on survival after BM than perhaps previously emphasized.

From our data analysis, it cannot be extracted that if the benefit of survival comes from intracranial or systemic effect. For most patients, complex complications resulted from both intracranial and extracranial progression, leads to patients’ death. Therefore, it is not clear whether the beneficial effect on survival comes from either of these two entities. We added a short discussion regarding this issue.

The meta-analyses from Petrelli and Lehrer are cited – these meta-analyses should be discussed together, not in separate paragraphs.  The authors should comment on what their current review/meta-analysis add to these three reviews, given that no studies but 1 is included from 2019 and to date in the analyses in this manuscript? Why is Rulli et al (Cancer 2019) not cited?  

Thank you for this comment. I edited the text so that all similar studies come together. Previous meta-analyses were smaller with fewer patients analyzed. For example, in Petrelli et al. 754 patients from 13 studies were finally selected for statistical analysis (fig 2 in their paper). About Rulli’s study, they had included target therapy as well as immunotherapy. I also included this study in our manuscript. Only Petrelli et al reported Hazard ratio which is compared to our study.

An increased risk of radionecrosis (RN) is one important side-effect that should be considered when combining intracranial RT and IT. I think this should be discussed further in the Discussion, and also considering the fact that RN may occur increasingly after 1-2 years and later, this should be discussed according to the follow-up times in papers cited in the Toxicities section. 125, 3776-3789, doi:10.1002/cncr.32375.

Thank to your nice comment, we added a relevant discussion to explain this important fact.

All included papers were retrospective. There are no published prospective or randomized studies on IT and BM RT, as far as I know. The authors should discuss further how this limits the results in their analyses, they only mention this in a summerical manner. What studies, in the opinion of the authors, should be conducted to further enlighten us on this topic, and they should provide an overview of examples of clinical trials that are underway/being conducted (from clinicaltrials.gov or other sources).  

As you mentioned and we have declared in limitation section, we could not find any prospective study on this subject. A search in “clinical trilas.gov” showed that 9 study are designed for this subject: two prospective cohorts and 7 randomized study. These studies are recruiting patients and we should look forward to see the result of these studies as well as other studies that are not registered in clinicaltrials.gov.

The authors should be more clear and specific as to what their meta-analaysis adds to those already published (see above). Also - What is the recommendation from the authors after performing this meta-analysis – the “take home message”? Should we strive to combine IT and RT for BM? And how? Or is the recommendation simply that further studies are needed? The latter we know – we do need further studies. But can this current analysis guide us while awaiting the results if such studies?  

Generally speaking, our study present the most up-to-date sum-up of available information about RT+IT in brain metastasis. it partially supports and adds to previous studies and suggests concurrent RT-IT for these patients. Dosage of radiotherapy and type of immunotherapy cannot be achieved from our study and needs other investigation. I re-wrote the conclusion section based on this statements.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

In my opinion, the authors did not address thoroughly most of my comments.  

Reviewer 3 Report

No further comments. Substantially improved version.