VEGF Expression in Colorectal Cancer Metastatic Lymph Nodes: Clinicopathological Correlation and Prognostic Significance
, Eduardo A. Garcia 4, Mesquita Rodrigues 1 and Adhemar Longatto 1,2,4,5Round 1
Reviewer 1 Report
The manuscript authored by Dr Ines Mazeda et al aims to investigate the status of some members of the VEGF family and their receptors (VEGF-A, VEGF-C, VEGF-D, VEGFR2 and VEGFR3) in lymph node metastases of human colorectal cancers, in relation with the expression of these factors in the primary tumor, clinicopathologic parameters, disease recurrence and overall survival. This immunohistological retrospective study reveals that VEGFR2 and VEGFR3 expression levels in lymph nodes are correlated with CRC relapse, and that VEGF-A expression level in primary tumor is related to VEGF-A, VEGF-C, VEGF-D, VEGFR2 and VEGFR3 immunostaining in lymph nodes. None of these markers was associated with patient survival. The study is well conducted and of importance considering the requirement to define more precise markers for the prognosis and the follow-up of patients with colorectal cancers.
In Table II, it is not clear for me to what correspond the percentage values, e.g. VEGF-A : total number of cases : 136 ; total number of cases with VEGF-A expression : 124 => expected percentage of positive tumors : 91%. Percentage mentioned in Table II : 59%. Does this value correspond to the percentage of immunoreactive cells in the tumors ? This would not be coherent with other Tables.
A scale bar should be included in the Figure 1
According to Table II, the total number of cases with VEGF-A expression is 124 ; the total number of cases with VEGFR-2 expression is 64. In Table VI, the total number of cases investigated for VEGFR-2 and VEGF-A expression is 48. Why IHC was not performed on the 64 samples ? Does it mean that 16 samples investigated for VEGFR-2 expression were not subjected to VEGF-A IHC ?
“Positive” under VEGF ligands (Table VI) is misleading and should be removed. Alternatively, “n” should be positioned closed to the p-value.
In the legend, it should be clearly mentioned : n positive (%): Total number of cases with expression of both VEGF ligand and the VEGF-R and respective percentage.
The Authors showed that the VEGF-A immunostaining in primary tumors is positively related to VEGF-A, VEGF-C, VEGF-D, VEGFR-2 and VEGFR-3 in the matched lymph node. They could also evaluate whether this is a peculiarity of lymph nodes or whether VEGF-A expression is correlated with VEGF-C, VEGF-D, VEGFR-2 and VEGFR-3 in the primary tumor.
To improve readability of Table VII, the Authors should replace "protein marker" in 1st column by "primary tumor”, and mention over the horizontal line "VEGF-A, VEGF-C, …" : “Metastatic lymph node”
It is not clear in the Material and Methods section what were the criteria of exclusion of the patients. Were the cancers of the 512 patients at all stages (I – IV) of the TNM classification ?
The Authors should mention in the Material and Methods section which type of adjuvant chemotherapies were applied. They should also mention the date of the end of the follow-up of the patients.
In Table I, the Authors mentioned 70 patients with metastasis at the time of diagnosis, but 50 patients with tumor stage IV. Do these metastases correspond to organ metastases, as well as subserosa deposit, mesenteric and pericolic tissue involvement ?
For many years the molecular typing of colorectal tumors (KRAS, BRAF, PIK3CA, MSI/MSS status) are routinely performed. If these informations are available, the Authors could include them in their study. This might provide a link between the expression of VEGFs and their receptors and the molecular alterations in these different colorectal cancer subtypes.
The Authors should provide the references of the antibodies used in the study. Abcam retails different anti-VEGF-A antibodies. It is worth noting the existence of different isoforms VEGF-A as a result of alternative splicing.
The Authors should carefully check for misprints, e.g. lines 126 and 149 : "VEGFR-D" -> "VEGF-D" ; line 257 "CRC and lymph node samples was collect" -> "CRC and lymph node samples were collected"; line 259 : "Data was" -> "Data were"
The Title of Tables IV and V is partially deleted.
Author Response
1. In Table II, it is not clear for me to what correspond the percentage values, e.g. VEGF-A: total number of cases: 136; total number of cases with VEGF-A expression : 124 => expected percentage of positive tumors : 91%. Percentage mentioned in Table II: 59%. Does this value correspond to the percentage of immunoreactive cells in the tumors? This would not be coherent with other Tables.
Response: We thank the reviewer for the comment. (n) is the total number of cases in which the immunoreaction “worked”; n positive (%) are the cases in which the immunoreaction was considered positive. For VEGF-A: total number of cases with immunoreaction: 136; total number of cases with VEGF-A positive expression: 124
2. A scale bar should be included in the Figure 1
Response: We thank the reviewer for the comment, but we do not know how to do it.
3. According to Table II, the total number of cases with VEGF-A expression is 124 ; the total number of cases with VEGFR-2 expression is 64. In Table VI, the total number of cases investigated for VEGFR-2 and VEGF-A expression is 48. Why IHC was not performed on the 64 samples? Does it mean that 16 samples investigated for VEGFR-2 expression were not subjected to VEGF-A IHC ?
Response: We thank the reviewer for the comment. Table VI crosses the simultaneous results for VEGF-A and VEGFR-2, meaning that 48 of the cases have the analysis of the 2 markers simultaneously.
4. “Positive” under VEGF ligands (Table VI) is misleading and should be removed. Alternatively, “n” should be positioned closed to the p-value. In the legend, it should be clearly mentioned: n positive (%): Total number of cases with expression of both VEGF ligand and the VEGF-R and respective percentage. Response: We thank the reviewer for the comment. All the suggestions were added in the table.
5. The Authors showed that the VEGF-A immunostaining in primary tumors is positively related to VEGF-A, VEGF-C, VEGF-D, VEGFR-2 and VEGFR-3 in the matched lymph node. They could also evaluate whether this is a peculiarity of lymph nodes or whether VEGF-A expression is correlated with VEGF-C, VEGF-D, VEGFR-2 and VEGFR-3 in the primary tumor.
Response: We thank the reviewer for the comment. In the primary CRC we observed that VEGF-C was correlated with VEGFR-3.
Martins S F, Garcia E A, Luz M A et al., “Clinicopathological correlation and prognostic significance of VEGF-A, VEGF-C, VEGFR-2 and VEGFR-3 expression in colorectal cancer,” Cancer Genomics Proteomics 10:55–67, 2013.
6. To improve readability of Table VII, the Authors should replace "protein marker" in 1st column by "primary tumor”, and mention over the horizontal line "VEGF-A, VEGF-C, …" : “Metastatic lymph node”
Response: We thank the reviewer for the comment. The alteration was realized.
7. It is not clear in the Material and Methods section what were the criteria of exclusion of the patients. Were the cancers of the 512 patients at all stages (I – IV) of the TNM classification?
Response: We thank the reviewer for the comment. The ppopulation was constituted by 512 patients submitted to curative surgical intervention by CRC between 1st January of 2005 and 1st January of 2010 (at all stages) but for this study the sample consist in 210 patients with CRC with regional lymph node metastasis (stage III and IV).
8. The Authors should mention in the Material and Methods section which type of adjuvant chemotherapies were applied. They should also mention the date of the end of the follow-up of the patients.
Response: We thank the reviewer for the comment. We have not the information of what regimen of chemotherapy was realized. The data of follow was introduced.
9. In Table I, the Authors mentioned 70 patients with metastasis at the time of diagnosis, but 50 patients with tumor stage IV. Do these metastases correspond to organ metastases, as well as subserosa deposit, mesenteric and pericolic tissue involvement ?
Response: We thank the reviewer for the comment. The 50 patients were patients with pathological comproved metastasis. The 70 patients were patients that at staging was suspected to have metastasis not confirmed at surgery, namely lymph node, mesenteric and pericolic implants
10. For many years the molecular typing of colorectal tumors (KRAS, BRAF, PIK3CA, MSI/MSS status) are routinely performed. If these informations are available, the Authors could include them in their study. This might provide a link between the expression of VEGFs and their receptors and the molecular alterations in these different colorectal cancer subtypes.
Response: We thank the reviewer for the comment. Tthe information was not introdutced because we don’t have this data.
11. The Authors should provide the references of the antibodies used in the study. Abcam retails different anti-VEGF-A antibodies. It is worth noting the existence of different isoforms VEGF-A as a result of alternative splicing.
Response: We don’t have this information.
12. The Authors should carefully check for misprints, e.g. lines 126 and 149 : "VEGFR-D" -> "VEGF-D" ; line 257 "CRC and lymph node samples was collect" -> "CRC and lymph node samples were collected"; line 259 : "Data was" -> "Data were"
Response: We thank the reviewer for the comment. The correction was realized.
13. The Title of Tables IV and V is partially deleted.
Response: We thank the reviewer for the comment. The correction was realized.
Reviewer 2 Report
In this study, the authors monitored VEGF and VEGF-R protein expression in colorectal cancer metastatic lymph nodes from 210 CRC patients (either under 45 years of age or above 45 years of age) with metastatic diseases and examined whether the expression levels of VEGF family members and VEGF-receptors (VEGFR1 and VEGFR2) correlate with the patients clinical pathology and prognosis. Based on the expression levels of VEGF and VEGF-R the authors determined whether the correlation was significant to inform clinic pathological and prognostic significance. They also examined whether VEGF expression associates with age and/or survival of the patients based on which to recommend VEGF expression status in lymph node metastatic (LNM) samples of CRC patients as prognostic markers. Using IHC techniques in TMAs they assessed the expression levels of various VEGF family members and various VEGF receptors. They employed extensive statistical techniques to analyze their data. Their results showed that VEGF family expression is increased in LNM samples. They are report positive association between expression of VEGFR-2 and VEGFR-3 in LNM CRC relapse and potential prognostic value. They also observed very interesting relationship between the expression of these proteins in primary tumors and metastatic tumors. For example, they observed VEGF-A, VEGF-C, VEGF-D, VEGFR-2 and VEGFR-3 levels in primary tumors to correlate well with their respective levels in LNM. Based on their overall findings, the authors concluded that expression of various VEGF family members and VEGF receptors increases in LNM. Furthermore, the authors concluded there is association between expression of VEGFR-2 and VEGFR-3 in lymph nodes with CRC relapse, thus revealing a very important negative impact on the worse prognosis of the patients. However, they did not see any overall significant impact on CRC patient survival.
The main objective of this study was important, the experimental designs, methodological and statistical approaches used in these were standard and acceptable. For most part, data their presentation in tables were acceptable, although could be improved. The study’s findings address some of the important and otherwise unclear issues in regarding the potential roles of VEGR family of cytokines and their receptors in CRC. In conclusion, this is an important study which have produced very interesting and vital outcomes that could impact on how CRC patients with metastatic diseases are monitored and perhaps managed. From these perspectives, the study results and conclusions are highly relevant and will contribute to the field, thus generating future expansion in this field. However, there are some weaknesses in the current form of this manuscript, which require the authors’ responses.
Concerns/Weaknesses
- Data in table II: Please, provide a brief interpretation of the data and give a take away message from this table. That will be vital for the readers.
- Figure 1: Please, provide a brief overall summary of the results in this figure. That will be helpful to readers
- Table IV: Edit the title of this table by removing “and” from the title. Also, provide a sentence describing the main summary of the results shown in this table.
- Table V: Edit the title of this table by removing “and” from the title
- Table VII: In this table the word “positive” is misspelled across and should be corrected.
- Table VIII: Under VEGF-C, you showed that 6 out of the 18 patients who were negative for VEGF-C died. That is 6 of the 18 and that translates to 33.33%. But you showed 40%. Please, correct the error or explain why 6 out 18 gave you 40%.
- There are few typo errors and/or grammatical errors in the manuscript that show be corrected. An example is seen in line 230 where the word “reinforces” should be “reinforce”.
- On line 154 subtitle 2.6: You stated “overall survival curves according to expression of VEGF-A, VEGF-C, VEGF-D, VEGFR-2 and VEGFR-3 in CRC lymph nodes”. However, you did not show the survival curves. Rather, you summarized the survival data in table VIII. Since you did not show the curves, perhaps, you should consider changing that subtitle under 2.6. An example could “Lack of relationship between overall CRC patient survival and expression of VEGF-A, VEGF-C, VEGF-D, VEGFR-2 and VEGFR-3 in CRC lymph nodes.”
- Since the VEGF and VEGFR protein levels did not affect the patient survival, some readers may question the main clinical value of monitoring the levels of VEGF family and their receptors in LNM samples. Therefore, in your discussion section, you could stress more on how such measurements could be useful in monitoring how CRC patients with metastatic lymph nodes respond to therapy.
Author Response
"Please see the attachment."
1. Data in table II: Please, provide a brief interpretation of the data and give a take away message from this table. That will be vital for the readers.
Response: We thank the reviewer for the comment. A brief interpretation was introduced (line 87-88)
2. Figure 1: Please, provide a brief overall summary of the results in this figure. That will be helpful to readers
Response: We thank the reviewer for the comment. A brief summary of the figure is in line 93-94
3. Table IV: Edit the title of this table by removing “and” from the title. Also, provide a sentence describing the main summary of the results shown in this table.
Response: We thank the reviewer for the comment. The title was edited and the sentence was introduced “When analysing the correlation between these markers with data from diagnosis/surgery, like patient gender and age, personal or familiar history and follow-up data like relapse and death no significant association was found, as observed in Table IV.”
4. Table V: Edit the title of this table by removing “and” from the title
Response: We thank the reviewer for the comment. The title was edited.
5. Table VII: In this table the word “positive” is misspelled across and should be corrected.
Response: We thank the reviewer for the comment. It has been corrected.
6. Table VIII: Under VEGF-C, you showed that 6 out of the 18 patients who were negative for VEGF-C died. That is 6 of the 18 and that translates to 33.33%. But you showed 40%. Please, correct the error or explain why 6 out 18 gave you 40%.
Response: We thank the reviewer for the comment. It has been corrected.
7. There are few typo errors and/or grammatical errors in the manuscript that show be corrected. An example is seen in line 230 where the word “reinforces” should be “reinforce”.
Response: We thank the reviewer for the comment. It has been corrected.
8. On line 154 subtitle 2.6: You stated “overall survival curves according to expression of VEGF-A, VEGF-C, VEGF-D, VEGFR-2 and VEGFR-3 in CRC lymph nodes”. However, you did not show the survival curves. Rather, you summarized the survival data in table VIII. Since you did not show the curves, perhaps, you should consider changing that subtitle under 2.6. An example could “Lack of relationship between overall CRC patient survival and expression of VEGF-A, VEGF-C, VEGF-D, VEGFR-2 and VEGFR-3 in CRC lymph nodes.”
Response: We thank the reviewer for the comment. It has been corrected.
9. Since the VEGF and VEGFR protein levels did not affect the patient survival, some readers may question the main clinical value of monitoring the levels of VEGF family and their receptors in LNM samples. Therefore, in your discussion section, you could stress more on how such measurements could be useful in monitoring how CRC patients with metastatic lymph nodes respond to therapy.
Response: We thank the reviewer for the comment. Despite the data obtained, they do not directly impact the outcome of patients
