New Genetic Technologies in Diagnosis and Treatment of Cancer of Unknown Primary

Round 1

Reviewer 1 Report

The authors have done a fairly good job of highlighting and describing the challenges of the management of patients with cancer of unknown primary (CUP). They presented a clear summary of the state of the art of CUP management and proposed future potential novel strategies for its improvement. The review is timely and will be of interest to the scientific community and the readers of Cancers journal.

I suggest the authors proofreading to check for structural errors in the sentences and typos.

Author Response

Answers to Reviewers comments:

On behalf of all authors, I thank the Reviewers for their encouraging feedback, valuable comments, and suggestions. We have carefully considered the Reviewers' comments and respond to them point-by-point as follows (in italics). All changes in the manuscript were made with “track changes” turned on.

Reviewer 1

The authors have done a fairly good job of highlighting and describing the challenges of the management of patients with cancer of unknown primary (CUP). They presented a clear summary of the state of the art of CUP management and proposed future potential novel strategies for its improvement. The review is timely and will be of interest to the scientific community and the readers of Cancers journal.

We thank the Reviewer for the positive appraisal of our article

I suggest the authors proofreading to check for structural errors in the sentences and typos.

The article has been subjected to thorough proofreading made by a professional company.

Reviewer 2 Report

it is an interesting and well presented review

more information regatrding the today clinical practice and how these technologies can lead the therapeutic procedure

Author Response

Reviewer 2

It is an interesting and well presented review.

More information regarding the today clinical practice and how these technologies can lead the therapeutic procedure.

We thank the Reviewer for the positive appraisal of our article. We have slightly modified paragraphs addressing current clinical practice to comply with the Reviewer’s suggestion.

Reviewer 3 Report

I concur with the authors that the role of molecular testing in CUP will (and should) increase, and that guidelines need updating incorporating the new molecular data. However, I do not concur with the authors taking the role of ToO for granted, because there is no randomized data to support this conclusion.

Major Points:

Introduction Line 45: “… and may not allow to determining the tissue of origin…”; this is a bit misleading; by the very nature of CUP, the primary is elusive. So immunohistochemistry might provide hints and only in a few cases a specific profile as is the case for colon-like-CUP, but in most cases the immunohistochemistry is offering circumstantial evidence. Therefore, the term ToO is usually not used in the context of immunohistochemistry, but is reserved to RNA based molecular techniques

Introduction. Line 48-49: not correct, meta-analyses of autopsy series have shown that lung and pancreas are most frequent. Admittedly, in molecular ToO tests the frequency of colorectal cancers is higher. So the authors should be more specific whether they provide autopsy or molecular ToO data.

Introduction. Line 54: please replace “generalized” by “metastatic”, because there are also patients with a single metastatic site

Introduction Line 55: Establishment of ToO may not be essential, the authors refer themselves later to the negative randomized trial. So this sentence should not be so apodictic.

Line 97: Table 1 is problematical and I would recommend to skip it. By the very nature of CUP, no reliable cross-validation of the ToO predictions is possible, and the detailed percentages in the Table therefore create a false illusion of accuracy. In studies, the soundness of predictions is merely counterchecked by clincial and (immuno)histological plausibility. Yes, the respective papers make the mistake of presenting these percentage numbers uncritically, but this should not be adapted in the review.

Figure 1: should be skipped: When there is not enough material for sound immunohistology, a rebiopsy should be sought. There is no evidence or data to support that a liquid biopsy can make up for material deficiency in immunohistochemistry. Also, there is no data that a treatment tailored to the presumptive primary is better than empiric chemotherapy. So this Figure oversimplifies things and is not covered by data. It should be skipped altogether. There is also no randomized evidence for superiority of ToO, which is presented here as standard.

Line 114/115: This is confusing to the reader. The authors should make a clear distinction between 1.) ToO based molecular essays (RNA or methylation essays to predict the primary) and 2.) NGS sequencing to detect targets for molecular therapy (and maybe a small chance for primary prediction when a classical driver alteration is found). But generally, these are two methodologically different approaches, and they should be clearly separated.   

Line 118 – 131 and 142-151: this paragraph could be shortened, since healthy controls are not so relevant to the topic of the review

Line 152 - 154: I would not deem NGS sequencing a technique able to assign a likely primary that often. The authors imply that at least a probable primary can be predicted. That is overoptimistic from NGS data. Note that all NGS sequencing studies have focused on frequencies of actionable alterations, but have refrained from frequencies of ToO prediction.

Line 204: The role of ToO is not settled yet. But randomized trials (both RNA based) were negative. The cited methylation based assay was not randomized, so data might be tainted with selection bias. In the whole, to my judgment the merits of ToO and the potential of NGS sequencing to contribute to ToO are overstated.  

Line 241: The frequency of mutations deemed targetable in NGS sequencing studies varies depending on the clinicians´assignment which mutations are targetable. The 47% is based on very optimistic assumptions where also cases with a weak molecular rationale and hardly any underlying clinical evidence are counted as well. In reality and practice, numbers are lower.

Line 348: “The rational management of CUP necessitates establishing ToO”. I do not concur. With obvious exceptions (renal CUP, colon-like CUP) this thesis has not been supported by any randomized data. Patients might suffer from a prolonged unsuccessful search for the primary and may be better served by a more rapid initiation of therapy.

Finally, numerous reviews have been written in the field of CUP covering the modern molecular techniques. There is a discrepancy between numerous reviews as opposed to fewer recent studies providing original research data. So a main concern with the manuscript is that any review in CUP written at this point of time covers aspects already previously reviewed. It is really hard to find a new perspective or new angle to review the data. This is also a problem with this manuscript.

Minor Points:

Abstract Line 14/15: rephrase sentence, is not correct grammar-wise  

Line 97: immunohistochemical instead of immunihistochemical

Author Response

Answers to Reviewers comments:

On behalf of all authors, I thank the Reviewers for their encouraging feedback, valuable comments, and suggestions. We have carefully considered the Reviewers' comments and respond to them point-by-point as follows (in italics). All changes in the manuscript were made with “track changes” turned on.

Reviewer 3

Major Points:

Introduction Line 45: “… and may not allow to determining the tissue of origin…”; this is a bit misleading; by the very nature of CUP, the primary is elusive. So immunohistochemistry might provide hints and only in a few cases a specific profile as is the case for colon-like-CUP, but in most cases, the immunohistochemistry is offering circumstantial evidence. Therefore, the term ToO is usually not used in the context of immunohistochemistry but is reserved to RNA based molecular techniques

In the current version, we rephrased this paragraph and put ToO in the context of RNA-based molecular techniques.

Introduction. Line 48-49: not correct, meta-analyses of autopsy series have shown that lung and pancreas are most frequent. Admittedly, in molecular ToO tests the frequency of colorectal cancers is higher. So the authors should be more specific whether they provide autopsy or molecular ToO data.

We corrected this paragraph accordingly to make it more precise.

Introduction. Line 54: please replace “generalized” by “metastatic”, because there are also patients with a single metastatic site

We replaced “generalized” with “metastatic”.

Introduction Line 55: Establishment of ToO may not be essential, the authors refer themselves later to the negative randomized trial. So this sentence should not be so apodictic.

We have replaced the phrase “is essential for planning appropriate treatment” with “may inform treatment selection”.

Line 97: Table 1 is problematical and I would recommend to skip it. By the very nature of CUP, no reliable cross-validation of the ToO predictions is possible, and the detailed percentages in the Table therefore create a false illusion of accuracy. In studies, the soundness of predictions is merely counterchecked by clinical and (immuno)histological plausibility. Yes, the respective papers make the mistake of presenting these percentage numbers uncritically, but this should not be adapted in the review.

In table 1 the “accuracy” has been replaced by “Reported concordance with a clinical and (immune)histological diagnosis” to avoid overinterpretation. We have also added the following comment: “Notably, the concordance percentages should be considered cautiously, as they lack cross-validation of the ToO predictions and counterchecking by clinical and immuno-histological plausibility”. This table provides a summary of the used tests and a general view of their potential applications. However, we leave the decision to keep or remove this table to the editors.

Figure 1: should be skipped: When there is not enough material for sound immunohistology, a rebiopsy should be sought. There is no evidence or data to support that a liquid biopsy can make up for material deficiency in immunohistochemistry. Also, there is no data that a treatment tailored to the presumptive primary is better than empiric chemotherapy. So this Figure oversimplifies things and is not covered by data. It should be skipped altogether. There is also no randomized evidence for superiority of ToO, which is presented here as standard.

We have deleted the Figure 1.

Line 114/115: This is confusing to the reader. The authors should make a clear distinction between 1.) ToO based molecular assays (RNA or methylation essays to predict the primary) and 2.) NGS sequencing to detect targets for molecular therapy (and maybe a small chance for primary prediction when a classical driver alteration is found). But generally, these are two methodologically different approaches, and they should be clearly separated.   

We made the proposed changes to clearly distinct these two approaches.

Line 118 – 131 and 142-151: this paragraph could be shortened, since healthy controls are not so relevant to the topic of the review

Both paragraphs were shortened, as suggested.

Line 152 - 154: I would not deem NGS sequencing a technique able to assign a likely primary that often. The authors imply that at least a probable primary can be predicted. That is overoptimistic from NGS data. Note that all NGS sequencing studies have focused on frequencies of actionable alterations, but have refrained from frequencies of ToO prediction.

The sentence has been nuanced to avoid an over-optimistic impression of the value of NGS to predict ToO.

Line 204: The role of ToO is not settled yet. But randomized trials (both RNA based) were negative. The cited methylation based assay was not randomized, so data might be tainted with selection bias. In the whole, to my judgment the merits of ToO and the potential of NGS sequencing to contribute to ToO are overstated.  

We explained at the end of this paragraph that in that study patients with different methylation genes status were not randomized to the study groups and that the observations may not be entirely reliable.

Line 241: The frequency of mutations deemed targetable in NGS sequencing studies varies depending on the clinicians´assignment which mutations are targetable. The 47% is based on very optimistic assumptions where also cases with a weak molecular rationale and hardly any underlying clinical evidence are counted as well. In reality and practice, numbers are lower.

The value of 47% was calculated based on data available in the literature. We agree with the Reviewer that the real-life values in clinical practice may be lower, and this was mentioned. Additionally, we now underlined the lack of a clear relationship between the presence of some molecular abnormalities and the effectiveness of targeted therapies in clinical trials.

Line 348: “The rational management of CUP necessitates establishing ToO”. I do not concur. With obvious exceptions (renal CUP, colon-like CUP) this thesis has not been supported by any randomized data. Patients might suffer from a prolonged unsuccessful search for the primary and may be better served by a more rapid initiation of therapy.

We softened the first sentence of the Conclusion and added information that prompt therapy should not be compromised by a prolonged ToO target searching.

There is a discrepancy between numerous reviews as opposed to fewer recent studies providing original research data. So a main concern with the manuscript is that any review in CUP written at this point of time covers aspects already previously reviewed.

We realize that there is a scarcity of original studies addressing this topic. However, in view of the rapid development of personalized cancer therapy, the role of this issue will certainly grow. “Cancers”, thanks to its open access format, has many readers and we hope that our work will attract their interest.

Minor Points:

Abstract Line 14/15: rephrase sentence, is not correct grammar-wise  

Line 97: immunohistochemical instead of immunohistochemical

All minor points were corrected.

Reviewer 4 Report

Dear authors and editor,

The manuscript titled ‘’ New genetic technologies in diagnosis and treatment of cancer of unknown primary’’ analyzes the problem of cancer of unknown primary. In clinical practice, this problem sometimes affects up to 3 percent of cases. As an oncologist surgeon, I affirm that this problem we can find most often in cancers of the stomach, breast, pancreas as well as cancers of the lung and colon, including the rectum. In such situations radical surgical treatment is not an option, only chemotherapy remains. The prognosis in such patients is usually bad, therefore individually selected chemotherapy appears to be the best treatment option. In my opinion the manuscript is an important significance for individualized treatment. At present, there are no recommendations for the rules of conduct in this type of situation.

The NGS technique is very good, sensitive and more and more enters in to the daily clinical practice. My research group is working on collecting genetic material obtained from a single cell and analysis using NGS. Congratulations on the topic and article.

The paper is well-organised, the language is correct and the content is understandable. Literature correctly selected.

In conclusion, I support publication of the presented article.

Thank you for your choice me as a reviewer.

Author Response

Reviewer 4

The manuscript titled ‘’New genetic technologies in diagnosis and treatment of cancer of unknown primary’’ analyzes the problem of cancer of unknown primary. In clinical practice, this problem sometimes affects up to 3 percent of cases. As an oncologist surgeon, I affirm that this problem we can find most often in cancers of the stomach, breast, pancreas as well as cancers of the lung and colon, including the rectum. In such situations radical surgical treatment is not an option, only chemotherapy remains. The prognosis in such patients is usually bad, therefore individually selected chemotherapy appears to be the best treatment option. In my opinion, the manuscript is an important significance for individualized treatment. At present, there are no recommendations for the rules of conduct in this type of situation.

The NGS technique is very good, sensitive and more and more enters into the daily clinical practice. My research group is working on collecting genetic material obtained from a single cell and analysis using NGS. Congratulations on the topic and article.

The paper is well-organised, the language is correct and the content is understandable. Literature correctly selected.

We thank the Reviewer for the positive appraisal of our article.

Round 2

Reviewer 3 Report

The frequencies of putative primary organs in autopsy and ToO should be backed up with citations.