Prognostic Implications of Intratumoral and Peritumoral Infiltrating Lymphocytes in Pancreatic Ductal Adenocarcinoma

Round 1

Reviewer 1 Report

There are some questions for authors: 

1. What does the “O” stand for in the table 1, the authors should explain it in the annotation.
2. Sadozai, 2021, the authors should explain where is “various” in Table 1
3. In table 1, why the number of the patients can’t correlate with tumor stage? There are 145 patients with tumor stage ranging from I to III in Diana 2016 et al and 108 patients with stage I-II in Tahkola et al 2018, but only 80 patients with stage ranging from I to IV in Fukunaga et al 2004. There is same situation in Mota Reyes 2019 et al and Seifert 2021 et al. The tumor staging is uneven in the article where the authors included. The authors should explain.
4. In material and methods, the authors use PubMed database as their searching website. The authors should also consider the articles from other database for a comprehensive meta-analysis article such as Embase, Web of science, Scopus.
5. In table 2, why the Egger’s tests for CD4 and FOXP3 are “-”?
6. Similarly, why the Egger’s test in table 3 are “-” in some tests?
7. The figure 2 should be clearer, especially Figure 2 A-C.
8. The authors should explain “The choice of immunotherapeutic agent may be important in pancreatic cancers because they can be found as higher TNM stages during diagnosis” in detailed on page 6 Lines 159-161.
9. Why the authors wrote “Because the evaluation criteria differ between studies, further cumulative studies are needed for a detailed analysis of the criteria for high TIL levels in PDAC.” as the conclusion? The authors should explain in detailed.
10. The author mentioned that “The difference in the prognostic impact of TILs between the whole tumor area and the hot spots remain unclear.” on page 8 lines 228-229. But they also mentioned “In conclusion, high TIL levels were significantly correlated with a better PDAC prognosis” on page 8 line 240-241. Are these two sentences reasonable?

Author Response

There are some questions for authors:

1. What does the “O” stand for in the table 1, the authors should explain it in the annotation.

Response:

As the reviewer’s recommendation, the analyzed parameters were directly described in Table 1 as below:

 

Table 1. Main characteristics of the eligible studies

Authors	Location	No of Patients	Tumor stage	No of Chemo- radiotherapy	Analyzed parameters
Diana 2016	Canada	145	I-III	126	CD8, FOXP3
Fukunaga 2004	Japan	80	I-IV	0	CD4, CD8
Homma 2014	Japan	22	I-III	17	CD8
Hwang 2016	Korea	30	I-III	0	CD4
Ino 2013	Japan	212	I-IV	94	CD4, CD8, FOXP3
Liu 2015	China	72	I-III	ND	CD8
Liu 2016	China	92	I-III	92	CD8
Michelakos 2020	USA	133	I-II	63	CD8
Mota Reyes 2019	Germany	74	I-IV	37	CD4
Nejati 2017	USA	136	I-IV	136	CD4
Sadozai 2021	Switzerland	112	I-III	ND	CD3
Seifert 2021	Germany	69	I-IV	17	CD3, CD4, CD8
Tahkola 2018	Finland	108	I-II	0	CD3, CD8, Immune cell score
Tahkola 2019	Finland	79	I-III	0	CD3, CD8, Immune cell score
Tang 2014	USA	228	I-IV	0	CD4, CD8
Tewari 2013	UK	81	I-III	ND	CD3, CD8
Tsukamoto 2019	Japan	235	I-IV	ND	CD8
Wang 2017	China	120	II	120	FOXP3
Wartenberg 2015	Switzerland	120	I-IV	120	CD8
Zhang 2019	China	305	I-III	0	CD3, CD8, FOXP3

 

 

2. Sadozai, 2021, the authors should explain where is “various” in Table 1

Response:

We corrected the study location from “various” to “Switzerland.”

 

3. In table 1, why the number of the patients can’t correlate with tumor stage? There are 145 patients with tumor stage ranging from I to III in Diana 2016 et al and 108 patients with stage I-II in Tahkola et al 2018, but only 80 patients with stage ranging from I to IV in Fukunaga et al 2004. There is same situation in Mota Reyes 2019 et al and Seifert 2021 et al. The tumor staging is uneven in the article where the authors included. The authors should explain.

Response:

As described, the patient’s number was not included in the inclusion and exclusion criteria. In addition, the tumor stage was not included. We added the comments for the exclusion criteria in the methods part.

 

4. In material and methods, the authors use PubMed database as their searching website. The authors should also consider the articles from other database for a comprehensive meta-analysis article such as Embase, Web of science, Scopus.

Response:

As the reviewer’s recommendation, we performed the additional searching in the other database (MEDLINE). We added the information for the additional searching in the revised manuscript.

 

5. In table 2, why the Egger’s tests for CD4 and FOXP3 are “-”?

Response:

The minimal requirement of subset or article number is three to perform Egger’s test. However, in the assessment of CD4 and FOXP3, the included numbers of subsets were 1 and 2, respectively. Therefore, Egger’s test could not be performed. To diminish the misunderstanding, we changed from “-” to “NA.” In addition, the full name of “NA” is inserted in the footnote of Table 2.

 

6. Similarly, why the Egger’s test in table 3 are “-” in some tests?

Response:

Egger’s test could not be performed due to too small number of subsets. To diminish the misunderstanding, we changed from “-” to “NA.” In addition, the full name of “NA” is inserted in the footnote of Table 2.

 

7. The figure 2 should be clearer, especially Figure 2 A-C.

Response:

To compare between overall and stage I-III subgroups, the result of figure 2 was combined to new Table 2 as below:

 

Table 2. Correlation between intratumoral infiltrating lymphocytes and overall survival in pancreatic ductal adenocarcinoma

	Number of subsets	Fixed effect [95% CI]	Heterogeneity test [P-value]	Random effect [95% CI]	Egger’s Test [P-value]
Intratumoral CD3	5	0.747 [0.620, 0.900]	0.714	0.747 [0.620, 0.900]	0.138
Stage I to III	4	0.721 [0.590-0.880]	0.778	0.721 [0.590-0.880]	0.303
Western population	4	0.826 [0.648, 1.052]	0.917	0.826 [0.648, 1.052]	0.945
Eastern population	1	0.646 [0.482, 0.865]	1.000	0.646 [0.482, 0.865]	NA
Intratumoral CD4	6	0.755 [0.632, 0.902]	0.481	0.755 [0.632, 0.902]	0.424
Stage I to III	1	0.618 [0.261, 1.463]	1.000	0.618 [0.261, 1.463]	NA
Western population	3	0.774 [0.602, 0.994]	0.127	0.745 [0.515, 1.079]	0.361
Eastern population	3	0.736 [0.572, 0.949]	0.865	0.736 [0.572, 0.949]	0.736
Intratumoral CD8	12	0.804 [0.711, 0.910]	0.003	0.754 [0.611, 0.930]	0.053
Stage I to III	8	0.833 [0.699, 0.994]	< 0.001	0.695 [0.478, 1.012]	0.035
Western population	5	0.786 [0.648, 0.953]	0.530	0.786 [0.648, 0.953]	0.452
Eastern population	7	0.817 [0.696, 0.959]	< 0.001	0.709 [0.496, 1.013]	0.151
Intratumoral FOXP3	4	1.363 [1.133, 1.640]	0.338	1.358 [1.115, 1.655]	0.307
Stage I to III	3	1.397 [1.112, 1.755]	0.198	1.361 [1.010, 1.835]	0.238
Western population	1	0.965 [0.561. 1.660]	1.000	0.965 [0.561. 1.660]	NA
Eastern population	3	1.426 [1.172, 1.737]	0.448	1.426 [1.172, 1.737]	0.973
Immune cell score	2	0.776 [0.566, 1.065]	< 0.001	0.776 [0.566, 1.065]	NA
Stage I to III	2	0.776 [0.566, 1.065]	< 0.001	0.776 [0.566, 1.065]	NA
Western population	2	0.776 [0.566, 1.065]	< 0.001	0.776 [0.566, 1.065]	NA
CI, Confidence interval; NA, Not applicable

 

So, to avoid duplication, figure 2 was deleted in the revised manuscript.

 

8. The authors should explain “The choice of immunotherapeutic agent may be important in pancreatic cancers because they can be found as higher TNM stages during diagnosis” in detailed on page 6 Lines 159-161.

Response:

To diminish the misunderstanding, we corrected the comment as below:

 

The importance of immunotherapy is increasing in various cancers. In addition, in pancreatic cancers, despite the lack of clinical efficacy, immunotherapeutic approaches have been performed for the treatment of PDAC patients [37].

 

Reference

37. Wandmacher, A.M.; Letsch, A.; Sebens, S. Challenges and Future Perspectives of Immunotherapy in Pancreatic Cancer. Cancers (Basel) 2021, 13, 4235.

 

9. Why the authors wrote “Because the evaluation criteria differ between studies, further cumulative studies are needed for a detailed analysis of the criteria for high TIL levels in PDAC.” as the conclusion? The authors should explain in detailed.

Response:

Evaluation criteria can affect the proportion of patients with high or low TIL rates. Therefore, the possibility of changing prognostic implication of TIL level according to the evaluation criteria is present. However, to diminish the misunderstanding, we deleted the comment.

 

10. The author mentioned that “The difference in the prognostic impact of TILs between the whole tumor area and the hot spots remain unclear.” on page 8 lines 228-229. But they also mentioned “In conclusion, high TIL levels were significantly correlated with a better PDAC prognosis” on page 8 line 240-241. Are these two sentences reasonable?

Response:

The extent of the evaluation area, whether the whole region or hot spot, can affect the proportion of patients with high or low TIL rate. Therefore, the possibility of changing prognostic implication of TIL level according to the evaluation method is present. However, to diminish the misunderstanding, we deleted the comment. In addition, the difference between studies with whole section and hot spot could not be found in this study. We added the comment for the limitation in the revised manuscript.

Author Response File: Author Response.docx

Reviewer 2 Report

It is the personal belief of this reviewer that meta-analysis obtained from studies published under different concepts and measuring different data and statistical methods do not provide a reliable result (see Wanous et al.[1]). In addition, different pathologists report their findings in a not uniform pattern and using different molecular techniques for cell identification that may confound the results.

Furthermore, four confounding factors can be identified in the manuscript:

• Most of the papers examined by the authors contained PDAC patients stage I-III, but there were many others with stage I to IV. Therefore, populations under comparison are not exactly the same. Stage IV represents a more evolved tumor and it is possible that they have a different T-cell profile. We do not know.
• Chronic pancreatitis is an important cause of PDAC. Pancreatic inflammation is severe and this elicits the presence of inflammatory and immunogenic cells like T-lymphocytes. There is no distinction in the analysis whether the patients had (or not) a history of pancreatitis. Those who had pancreatitis require a separate sub-grouping.
• There are no data on previous chemotherapy which can substantially modify the type of T-cell infiltration.
• The studies analyzed by the authors contain different ethnic groups (see Table 1). Western populations (11 papers) show a tendency to a higher pancreatitis rate in their history than Eastern populations (9 papers).

 

Leaving aside this “personal belief” and confounding factors, the paper under scrutiny is:

• a well written manuscript;
• English is acceptably good;
• There are some items that require further clarification as follows:

 

Line 139 Generally, pancreatic cancers have worse survival rates, at only 7% to 8% [35].

This concept should be better explained. I suppose that they are talking of 5 year survival rate.

 

Line 141 Immunotherapy may be important in the treatment of advanced pancreatic cancers.

This requires a reference.

 

Line 156 In preoperative evaluations, chronic inflammation of the pancreas can be confused as a malignancy.

This requires a reference.

 

Line 163 the correlation between PD-L1 and TILs can be important in predicting the immunotherapeutic effect to and the prognosis of PDACs.

Besides the possible typo, this seems speculative.

 

Line 169 Higher CD8-expressing T lymphocyte density in tumor tissue was significantly correlated with the expression of cytotoxicity genes in pancreatic cancers.

It should say …in pancreatic cancer’s T lymphocytes.

 

 

 

[1] Wanous, J. P., Sullivan, S. E., & Malinak, J. (1989). The role of judgment calls in meta-analysis. Journal of Applied Psychology, 74(2), 259.

Author Response

It is the personal belief of this reviewer that meta-analysis obtained from studies published under different concepts and measuring different data and statistical methods do not provide a reliable result (see Wanous et al.[1]). In addition, different pathologists report their findings in a not uniform pattern and using different molecular techniques for cell identification that may confound the results.

 

Furthermore, four confounding factors can be identified in the manuscript:

 

Most of the papers examined by the authors contained PDAC patients stage I-III, but there were many others with stage I to IV. Therefore, populations under comparison are not exactly the same. Stage IV represents a more evolved tumor and it is possible that they have a different T-cell profile. We do not know.

Response:

As a comment, each study included various populations. We performed the additional analysis for patients with stage I to III (Supplementary Table 1).

 

Supplementary Table 1. Correlation between intratumoral infiltrating lymphocytes and overall survival in pancreatic ductal adenocarcinoma according to the tumor stage (I-III)
	Number of subsets	Fixed effect [95% CI]	Heterogeneity test [P-value]	Random effect [95% CI]	Egger’s Test [P-value]
CD3
Intratumoral high vs. low	4	0.721 [0.590-0.880]	0.778	0.721 [0.590-0.880]	0.303
CD4
Intratumoral high vs. low	1	0.618 [0.261, 1.463]	1.000	0.618 [0.261, 1.463]	NA
CD8
Intratumoral high vs. low	8	0.833 [0.699, 0.994]	< 0.001	0.695 [0.478, 1.012]	0.035
FOXP3
Intratumoral high vs. low	3	1.397 [1.112, 1.755]	0.198	1.361 [1.010, 1.835]	0.238
Immune cell score
Intratumoral high vs. low	2	0.776 [0.566, 1.065]	< 0.001	0.776 [0.566, 1.065]	NA
CI, Confidence interval

 

In the revised manuscript, this result was included.

 

 

Chronic pancreatitis is an important cause of PDAC. Pancreatic inflammation is severe and this elicits the presence of inflammatory and immunogenic cells like T-lymphocytes. There is no distinction in the analysis whether the patients had (or not) a history of pancreatitis. Those who had pancreatitis require a separate sub-grouping.

Response:

As pointed out, in most cases, cancer tissue might be accompanied by chronic pancreatitis [15]. We investigated the number or history of pancreatitis as below:

 

Authors	Location	No of Patients	No of Pancreatitis	Tumor stage	No of Chemo- radiotherapy	Analzed parameters
Diana 2016	Canada	145	ND	I-III	126	CD8, FOXP3
Fukunaga 2004	Japan	80	Most cases	I-IV	0	CD4, CD8
Homma 2014	Japan	22	ND	I-III	17	CD8
Hwang 2016	Korea	30	ND	I-III	0	CD4
Ino 2013	Japan	212	ND	I-IV	94	CD4, CD8, FOXP3
Liu 2015	China	72	ND	I-III	ND	CD8
Liu 2016	China	92	ND	I-III	92	CD8
Michelakos 2020	USA	133	ND	I-II	63	CD8
Mota Reyes 2019	Germany	74	ND	I-IV	37	CD4
Nejati 2017	USA	136	ND	I-IV	136	CD4
Sadozai 2021	Switzerland	112	ND	I-III	ND	CD3
Seifert 2021	Germany	69	ND	I-IV	17	CD3, CD4, CD8
Tahkola 2018	Finland	108	ND	I-II	0	CD3, CD8, Immune cell score
Tahkola 2019	Finland	79	ND	I-III	0	CD3, CD8, Immune cell score
Tang 2014	USA	228	ND	I-IV	0	CD4, CD8
Tewari 2013	UK	81	ND	I-III	ND	CD3, CD8
Tsukamoto 2019	Japan	235	ND	I-IV	ND	CD8
Wang 2017	China	120	ND	II	120	FOXP3
Wartenberg 2015	Switzerland	120	ND	I-IV	120	CD8
Zhang 2019	China	305	ND	I-III	0	CD3, CD8, FOXP3

 

However, subgroup analysis could not performed due to insufficient information of pancreatitis. We added the comment in the limitation part of revised manuscript.

 

 

There are no data on previous chemotherapy which can substantially modify the type of T-cell infiltration.

Response:

As a recommendation, the change of the T-cell infiltration can be possible after chemotherapy. However, our study could not be performed due to insufficient information of eligible studies. We added the comment in the limitation part of the manuscript as below:

 

In addition, the change of TILs after chemotherapy could not be obtained.

 

 

The studies analyzed by the authors contain different ethnic groups (see Table 1). Western populations (11 papers) show a tendency to a higher pancreatitis rate in their history than Eastern populations (9 papers).

Response:

As a recommendation, we compared between Western and Eastern populations. In an additional analysis, there were different results in intratumoral CD4, CD8, and FOXP3 from the overall population (Red color in supplementary Table 2).

 

Supplementary Table 2. Correlation between intratumoral infiltrating lymphocytes and overall survival in pancreatic ductal adenocarcinoma.
	Number of subsets	Fixed effect [95% CI]	Heterogeneity test [P-value]	Random effect [95% CI]	Egger’s Test [P-value]
Intratumoral CD3	5	0.747 [0.620, 0.900]	0.714	0.747 [0.620, 0.900]	0.138
Western population	4	0.826 [0.648, 1.052]	0.917	0.826 [0.648, 1.052]	0.945
Eastern population	1	0.646 [0.482, 0.865]	1.000	0.646 [0.482, 0.865]	NA
Intratumoral CD4	6	0.755 [0.632, 0.902]	0.481	0.755 [0.632, 0.902]	0.424
Western population	3	0.774 [0.602, 0.994]	0.127	0.745 [0.515, 1.079]	0.361
Eastern population	3	0.736 [0.572, 0.949]	0.865	0.736 [0.572, 0.949]	0.736
Intratumoral CD8	12	0.804 [0.711, 0.910]	0.003	0.754 [0.611, 0.930]	0.053
Western population	5	0.786 [0.648, 0.953]	0.530	0.786 [0.648, 0.953]	0.452
Eastern population	7	0.817 [0.696, 0.959]	< 0.001	0.709 [0.496, 1.013]	0.151
Intratumoral FOXP3	4	1.363 [1.133, 1.640]	0.338	1.358 [1.115, 1.655]	0.307
Western population	1	0.965 [0.561. 1.660]	1.000	0.965 [0.561. 1.660]	NA
Eastern population	3	1.426 [1.172, 1.737]	0.448	1.426 [1.172, 1.737]	0.973
Immune cell score	2	0.776 [0.566, 1.065]	< 0.001	0.776 [0.566, 1.065]	NA
Western population	2	0.776 [0.566, 1.065]	< 0.001	0.776 [0.566, 1.065]	NA
Eastern population	0
CI, Confidence interval

 

We added this result in the revised manuscript.  

 

 

Leaving aside this “personal belief” and confounding factors, the paper under scrutiny is:

a well written manuscript;

English is acceptably good;

 

There are some items that require further clarification as follows:

Line 139 Generally, pancreatic cancers have worse survival rates, at only 7% to 8% [35].

Response:

To diminish the misunderstanding, we corrected the sentence as below:

 

Generally, pancreatic cancers have worse survival rates [35].

 

 

This concept should be better explained. I suppose that they are talking of 5 year survival rate.

Response:

Each study showed various follow-up periods for survivals. However, the follow-up period can affect the prognostic implication. We analyzed using five years survival rates. So, articles with a short follow-up period were excluded. This explanation added the exclusion criteria.

 

 

Line 141 Immunotherapy may be important in the treatment of advanced pancreatic cancers.

This requires a reference.

Response:

Because many trials of immunotherapeutic regimens have been performed in PDACs, the importance of immunotherapy is gradually increasing. To diminish the misunderstanding, we corrected the sentence and added the reference in the revised manuscript as below:

 

The trials for immunotherapeutic treatment of PDACs have been performed to improve the patients’ survival [37].

 

Reference

37. Wandmacher, A.M.; Letsch, A.; Sebens, S. Challenges and Future Perspectives of Immunotherapy in Pancreatic Cancer. Cancers (Basel) 2021, 13, 4235.

 

 

Line 156 In preoperative evaluations, chronic inflammation of the pancreas can be confused as a malignancy.

This requires a reference.

Response:

We added the reference in the revised manuscript.

 

Reference

Munigala, S.; Kanwal, F.; Xian, H.; Agarwal, B. New diagnosis of chronic pancreatitis: risk of missing an underlying pancreatic cancer. Am. J. Gastroenterol. 2014, 109, 1824-1830.

 

 

Line 163 the correlation between PD-L1 and TILs can be important in predicting the immunotherapeutic effect to and the prognosis of PDACs.

Besides the possible typo, this seems speculative.

Response:

As the recommendation, we deleted the sentence and added the comment as below:

 

However, in each cancer type, the implications of TILs may be different. PD-1 binding to PD-L1 involves tumor growth and progression by suppressing the an-ti-tumor activity of T cells [41]. After antigen recognition and activation, PD-1 is upregu-lated in T cells [41]. Hence, the correlation between PD-L1 and TILs can be important in predicting the immunotherapeutic effect to and the prognosis of PDACs. Therefore, the study is needed to improve the understanding of the implication of TILs in PDACs.

 

 

Line 169 Higher CD8-expressing T lymphocyte density in tumor tissue was significantly correlated with the expression of cytotoxicity genes in pancreatic cancers.

It should say …in pancreatic cancer’s T lymphocytes.

Response:

As a recommendation, we corrected the sentence as below:

 

Higher CD8-expressing T lymphocyte density in tumor tissue was significantly correlated with the expression of cytotoxicity genes in pancreatic cancers of T lymphocytes of pancreatic cancer.

Author Response File: Author Response.docx

Reviewer 3 Report

In the manuscript entitled, “Prognostic implications of intratumoral and peritumoral infiltrating lymphocytes in pancreatic ductal adenocarcinoma” by Pyo et al., where the authors have done meta-analysis to elucidate the prognostic significance of intratumoral and peritumoral infiltrating T-lymphocytes in pancreatic ductal adenocarcinoma (PDAC), there are no major concerns that have been observed in the manuscript that require attention. The authors have performed the studies in a simplistic manner and the description of the methods is clear. They have also addressed the shortcomings of the study. The manuscript is recommended for publication with improvement in the quality of the figures presented in the manuscript.

Author Response

In the manuscript entitled, “Prognostic implications of intratumoral and peritumoral infiltrating lymphocytes in pancreatic ductal adenocarcinoma” by Pyo et al., where the authors have done meta-analysis to elucidate the prognostic significance of intratumoral and peritumoral infiltrating T-lymphocytes in pancreatic ductal adenocarcinoma (PDAC), there are no major concerns that have been observed in the manuscript that require attention. The authors have performed the studies in a simplistic manner and the description of the methods is clear. They have also addressed the shortcomings of the study. The manuscript is recommended for publication with improvement in the quality of the figures presented in the manuscript.

Response:

To compare between overall and stage I-III subgroups, the table was newly included. In addition, the result of figure 2 was combined with new Table 2. In addition, figure 2 was deleted in the revised manuscript.

 

Table 2. Correlation between intratumoral infiltrating lymphocytes and overall survival in pancreatic ductal adenocarcinoma

	Number of subsets	Fixed effect [95% CI]	Heterogeneity test [P-value]	Random effect [95% CI]	Egger’s Test [P-value]
Intratumoral CD3	5	0.747 [0.620, 0.900]	0.714	0.747 [0.620, 0.900]	0.138
Stage I to III	4	0.721 [0.590-0.880]	0.778	0.721 [0.590-0.880]	0.303
Western population	4	0.826 [0.648, 1.052]	0.917	0.826 [0.648, 1.052]	0.945
Eastern population	1	0.646 [0.482, 0.865]	1.000	0.646 [0.482, 0.865]	NA
Intratumoral CD4	6	0.755 [0.632, 0.902]	0.481	0.755 [0.632, 0.902]	0.424
Stage I to III	1	0.618 [0.261, 1.463]	1.000	0.618 [0.261, 1.463]	NA
Western population	3	0.774 [0.602, 0.994]	0.127	0.745 [0.515, 1.079]	0.361
Eastern population	3	0.736 [0.572, 0.949]	0.865	0.736 [0.572, 0.949]	0.736
Intratumoral CD8	12	0.804 [0.711, 0.910]	0.003	0.754 [0.611, 0.930]	0.053
Stage I to III	8	0.833 [0.699, 0.994]	< 0.001	0.695 [0.478, 1.012]	0.035
Western population	5	0.786 [0.648, 0.953]	0.530	0.786 [0.648, 0.953]	0.452
Eastern population	7	0.817 [0.696, 0.959]	< 0.001	0.709 [0.496, 1.013]	0.151
Intratumoral FOXP3	4	1.363 [1.133, 1.640]	0.338	1.358 [1.115, 1.655]	0.307
Stage I to III	3	1.397 [1.112, 1.755]	0.198	1.361 [1.010, 1.835]	0.238
Western population	1	0.965 [0.561. 1.660]	1.000	0.965 [0.561. 1.660]	NA
Eastern population	3	1.426 [1.172, 1.737]	0.448	1.426 [1.172, 1.737]	0.973
Immune cell score	2	0.776 [0.566, 1.065]	< 0.001	0.776 [0.566, 1.065]	NA
Stage I to III	2	0.776 [0.566, 1.065]	< 0.001	0.776 [0.566, 1.065]	NA
Western population	2	0.776 [0.566, 1.065]	< 0.001	0.776 [0.566, 1.065]	NA
CI, Confidence interval; NA, Not applicable

 

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

I have no further request and comment. 

Congratulations