Title: Immunohistochemical staining of radixin and moesin in prostatic adenocarcinoma

Transcription

1 Author's response to reviews Title: Immunohistochemical staining of radixin and moesin in prostatic adenocarcinoma Authors: Tanner L Bartholow (bartholow.tanner@medstudent.pitt.edu) Uma R Chandran (chandranur@msx.upmc.edu) Michael J Becich (becich@pitt.edu) Anil V Parwani (parwaniav@upmc.edu) Version: 2 Date: 17 December 2010 Author's response to reviews: see over

2 Dear Reviewers, Thank you very much for your suggestions to improve our paper "Immunohistochemical staining of radixin and moesin in prostatic adenocarcinoma". Based on your comments we have made the following changes described below, listed by individual reviewer. Dr. Roberta Mazzucchelli 1. The authors could explain why in the two groups (one for radixin and one for moesin) there are not the same number of the tissue samples. We now explain this in Line in the Methods section. "While every effort was made to include all of the cases in this study, processing artifacts within some of the TMAs cores made them unscorable, hence resulting in a difference in the numbers of cases between the two stains." 2. In the section of material and methods the authors should describe where is the staining of the cells, in fact it not clearly specified. We now explain this in Line in the Methods section. "Staining for both radixin and moesin was assessed in the cytoplasm of the cells of the prostatic glandular epithelium." 3. Why the authors used different not parametric tests (One-Way ANOVA for Radixin and Kruskal-Wallis for Moesin) to compare the different groups of tissue. We now explain this in Line in the Methods section. "Two different statistical tests were used in this study, depending on whether the data in specific comparison met the statistical assumptions of being normally distributed and having equal variances between the groups, both necessary to perform a parametric analysis. If not, this necessitated a non-parametric Kruskal-Wallis test.

3 4. When the authors used the Stage of carcinoma they should specify if they have used the TNM or not; and if the used tissue was obtained from radical prostatectomy or from other surgical specimens because they use stage less the 2 and this is impossible in the case of radical prostatectomy. We now explain this in Lines and Lines in the Methods section. All specimens were originally obtained through either a radical prostatectomy, transurethral resection of the prostate, or via a needle biopsy of the prostate, with the majority obtained using the first two methods. The Clinical TNM, as opposed to the Pathologic TNM, staging classification was used to assess the specimens. 5. The authors could expand the discussion and/or conclusions to explain the differences that they found between PIN (pre-neoplastic lesions) and the carcinoma. We now clarify our Discussion and Conclusions to reflect this. In Lines in the Discussion section we now state There are several possible reasons for our finding that specimens of carcinoma have less average staining than HGPIN. This may represent the unique physiological progression of radixin from the pre-neoplastic state to the neoplastic state. Another possibility is that within specimens of HGPIN, there can be a wide spectrum of histologic findings. More specifically, this means that in addition to glands demonstrating HGPIN, there may also be some elements of coaccompanying normal histological architecture, which may assist in imparting a higher staining score on these specimens. This may be true, as previous work has shown radixin to be down regulated in some instances of lungs cancer in comparison to non-tumor lung tissue [12]. The finding that the difference between NDP and PCa specimens was not significant, then, may be a reflection of the smaller sample size of NDPs available for study. As there were fewer NDPs, the natural baseline variability

4 among their expression levels may have had a greater impact in precluding statistical significance despite the absolute staining of NDP being higher than PCa. In Lines of the Conclusions, we now state More specifically, given that a difference was observed between HGPIN and PCa, this may indicate that radixin has the potential to be a clinically useful biomarker, but larger studies still need to be conducted before any definitive conclusions can be made. Future studies could also look at the expression of radixin in specimens of metastatic prostatic adenocarcinoma in order to determine if radixin is a clinically useful marker to predict the risk of metastasis. 6. They also could add a reference about Merlin (Curr Protein Pept Sci September 1; 11(6): ) and the authors should briefly explain what is Merlin. We have added this reference and expand our discussion of Merlin. In Line of our conclusion, we now state "Merlin (moesin-ezrin-radixin-like protein), the product of the neurofibromatosis type 2 (NF2) gene, is another member of the Protein 4.1 superfamily that has an established function as a schwannoma and meningioma tumor suppressor protein, with NF2 mutations also seen in cases of thyroid cancer, mesothelioma, and melanoma [21]. Conversely, it has been proposed to bind CD44 when not phosphorylated, suppressing growth, and is believed to be active in states of high cell densities. 7. The Figure 7 has not a good quality. The authors described in B a diffuse staining in the cytoplasm that is not evident. It could be useful to make a panel with PIN, normal and carcinoma We have now modified this figure and created a Figure 8 with photomicrographs which better depict the staining.

5 Dr. Lai Fong Kok 1. According to the WHO blue book (Pathology and Genetics of Tumours of the Urinary System and Male Genital Organs), PIN lesion can be further divided into 2 groups, low and high grade PIN. Therefore, in the study it is better to state the grades of PIN in order to provide more information. We have now clarified this in Line of the Introduction to state "Low grade prostatic intraepithelial neoplasia was not studied both because the diagnosis is subjective and because it lacks clinical relevance. In general, the percentage of cases of HGPIN that feature prostatic adenocarcinoma on rebiopsy is 30% [18]. No specimens of HGPIN included in this study were diagnosed at the time as containing PCa." We have also modified PIN to HGPIN (high-grade PIN) throughout the paper to reflect this. 2. Based on the result, the staining scores for redixin for PIN is higher then NAC and PCa. In the discussion, however, the authors didn't discus or explain the reasons for high staining scores for PIN. Therefore, discussion and result are not correlated to each other and the discussion is not fully supported by the result. We now clarify our Discussion to reflect this. In Lines in the Discussion section we now state There are several possible reasons for our finding that specimens of carcinoma have less average staining than HGPIN. This may represent the unique physiological progression of radixin from the pre-neoplastic state to the neoplastic state. Another possibility is that within specimens of HGPIN, there can be a wide spectrum of histologic findings. More specifically, this means that in addition to glands demonstrating HGPIN, there may also be some elements of coaccompanying normal histological architecture, which may assist in imparting a higher staining score on these specimens. This may be true, as previous work has shown radixin to be down regulated in some instances of lungs cancer in comparison

6 to non-tumor lung tissue [12]. The finding that the difference between NDP and PCa specimens was not significant, then, may be a reflection of the smaller sample size of NDPs available for study. As there were fewer NDPs, the natural baseline variability among their expression levels may have had a greater impact in precluding statistical significance despite the absolute staining of NDP being higher than PCa. 3. Finally, the conclusion is not clearly/directly even staining for redixin and moesin are not possible for clinical practice. We were not entirely sure that we understood this suggestion, but we took it to mean that we needed to better discuss the clinical implications for the differences that we observed in the radixin staining and moesin staining. In Lines of the Conclusions, we now state More specifically, given that a difference was observed between HGPIN and PCa, this may indicate that radixin has the potential to be a clinically useful biomarker, but larger studies still need to be conducted before any definitive conclusions can be made. Future studies could also look at the expression of radixin in specimens of metastatic prostatic adenocarcinoma in order to determine if radixin is a clinically useful marker to predict the risk of metastasis. In Line of the Discussion, we now state The finding that moesin appears to be down regulated from Stage 2 to Stage 4 may seem counterintuitive, as moesin-positive tumors have been shown to demonstrate higher perineural invasion rates in pancreatic adenocarcinoma [17]. However, moesin expression patterns can vary by cancer type as moesin, like radixin, has been shown to be down regulated in cases of lung cancer [12]. Moreover, it is possible that this finding may reflect a late stage change in tumor physiology.

7 In Line of the Conclusions, we now state "While a significant decrease in moesin staining was noted between Stage 2 and Stage 4 prostatic adenocarcinomas, the actual staining intensities were close in absolute terms (Fig. 6). While this difference may reflect a change in physiology in the tissues between stages, additional larger studies will need to be conducted to determine its ability to correlate with stage prior to any clinical implementation. 4. The high PIN lesion is more prevalent in prostatic carcinoma. It is better to declare whether any PIN lesion is associated with PCa. We have now clarified this in Line of the Introduction to state We have now clarified this in Line of the Methods to state "Low grade prostatic intraepithelial neoplasia was not studied both because the diagnosis is subjective and because it lacks clinical relevance. In general, the percentage of cases of HGPIN that feature prostatic adenocarcinoma on rebiopsy is 30% [18]. No specimens of HGPIN included in this study were diagnosed at the time as containing PCa." 5. More personal data for the tissue donor should be provided, such as average age. In Lines of the Results, we now list the ages and standard deviations of the patients by tissue category. The average ages and standard deviations of the patients whose specimens were included in the TMAs at the time of specimen retrieval were for PCa, for BPH, for NDP, for NAC, and for HGPIN. 6. For the figure 1, it is better to put a marker between the groups which showed significant difference.

8 We have now modified our Figure 1 to reflect this. General Comments: 1. Please also do the following: (1) You state that you used specimens located in the Health Sciences Tissue Bank at the University of Pittsburgh Medical Center; please document, within your manuscript, who granted approval for the use of these specimens. In Lines of the Methods we now discuss this. This study received exempt approval (PRO ) from the University of Pittsburgh Institutional Review Board. This approval and abiding by the guidelines for usage of specimens from the Health Sciences Tissue Bank at the University of Pittsburgh permitted the use of all specimens included in this study. 2. Please also highlight (with 'tracked changes'/coloured/underlines/highlighted text) all changes made when revising the manuscript to make it easier for the Editors to give you a prompt decision on your manuscript. We have complied with this. 3. Ensure that your revised manuscript conforms to the journal style ( ). It is important that your files are correctly formatted. We have complied with the submission guidelines. In closing, thank you again for your suggestions and consideration of our manuscript. We look forward to hearing from you.

9 Sincerely, Tanner Bartholow