Comparison of GATA3, GCDFP15, Mammaglobin and SOX10 Immunocytochemistry in Aspirates of Metastatic Breast Cancer

Round 1

Reviewer 1 Report

This study evaluated the sensitivity of ER, GATA3, GCDFP15, MMG, and SOX10 immunohistochemistry (IHC) on 115 breast cancer fine-needle aspiration cell block specimens. My suggestions are as follows.

1. I find table 2 a bit difficult to read. I believe that it should be "'≥' 5%, moderate/strong staining intensity (+/-)". Please change the table title to "The expression of GATA3, gross cystic disease fluid protein-15 (GCDFP15), mammaglobin (MMG), and SOX10 in aspirates of metastatic breast cancer". Within the table, please remove "a)" and replace "b)" with something like "The proportion of positive tumor cells (%)". Please define which cutoff you use for the proportion (ex: ≥ 1%, any intensity).

2. To determine the lineage of a metastatic lesion, specificity is equally important as sensitivity. These IHC markers may also be expressed in other non-breast cancers and this should be discussed in the discussion & conclusion sections. 

3. Some editing marks should be corrected (ex: page 2, lines 17 & 18; highlights in table 5).

Author Response

We thank the review for his/her constructive comments.

1. I find table 2 a bit difficult to read. I believe that it should be "'≥' 5%, moderate/strong staining intensity (+/-)". Please change the table title to "The expression of GATA3, gross cystic disease fluid protein-15 (GCDFP15), mammaglobin (MMG), and SOX10 in aspirates of metastatic breast cancer". Within the table, please remove "a)" and replace "b)" with something like "The proportion of positive tumor cells (%)". Please define which cutoff you use for the proportion (ex: ≥ 1%, any intensity).

Table 2 has been revised with the legends changed, the missing ≥ added and the intensity level for mean and median level added

2. To determine the lineage of a metastatic lesion, specificity is equally important as sensitivity. These IHC markers may also be expressed in other non-breast cancers and this should be discussed in the discussion & conclusion sections.

In the discussion, relevant text was added “However, neither GATA3, GCDFP15, MMG nor SOX10 is entirely specific for breast can-cers. Hence, the role of GCDFP15 in increasing specificity of an immunostain panel, or for resolving cases of conflicting/equivocal GATA3, MMG or SOX10 staining cannot be disregarded”

3. Some editing marks should be corrected (ex: page 2, lines 17 & 18; highlights in table 5).

The editing marks are removed, and we apologize for the oversight.

Reviewer 2 Report

The study by the Tse team describes the possibility of using GATA3, GCDFP15, MMG, and SOX10 as diagnosis markers for metastatic breast cancer. After analyzing over 100 cases from real patients, the authors built a correlation network between the expression of  GATA3, GCDFP15, MMG, and SOX10 and ER-positive/negative breast cancers. The manuscript includes convincing statistics data and immunocytochemistry images and is of good logic.  In all, I recommend this manuscript be accepted by the Journal of Molecular Pathology with some minor revisions.

1. The color in Figure 1 is not fully convincing. The red and green blocks separately represent specific markers' positive and negative expressions. But some white blocks occur in the figure without any explanation. Some clarification is required. 
2. The number in Table 3 is confusing. In line 1 a), all is 6, but ER- plus ER+ is 5, which is not equal to all.
3. A scale bar should be presented in Figure 2 and 3.
4. Some revision trace remains in the manuscript: page 2, 2.2, lines 3 and 4; Some highlight squares in Table 5.

Author Response

1. The color in Figure 1 is not fully convincing. The red and green blocks separately represent specific markers' positive and negative expressions. But some white blocks occur in the figure without any explanation. Some clarification is required.

The white blocks are a result of the lines separating each transparent block. The figure was revised, and the white lines are removed.

2. The number in Table 3 is confusing. In line 1 a), all is 6, but ER- plus ER+ is 5, which is not equal to all.

In some cases ER was not assessable due to exhaustion of tissue, thus the totals from the ER- and ER+ categories do not always sum up to the figure under all.

Legends were added to explain the issue under the table.

3. A scale bar should be presented in Figure 2 and 3.

Scale bars were added to all figures

4. Some revision trace remains in the manuscript: page 2, 2.2, lines 3 and 4; Some highlight squares in Table 5.

The editing marks are removed, and we apologize for the oversight.