Author's response to reviews Title:Targeted deep sequencing of mucinous ovarian tumors reveals multiple overlapping RAS-pathway activating mutations in borderline and cancerous neoplasms. Authors: Robertson T Mackenzie (rmackenzie@bccrc.ca) Stefan Kommoss (stefan.kommoss@med.uni-tuebingen.de) Boris J Winterhoff (bwinterhoff@hotmail.com) Benjamin R Kipp (kipp.benjamin@mayo.edu) Joaquin J Garcia (garcia.joaquin@mayo.edu) Jesse Voss (voss.jesse@mayo.edu) Kevin Halling (halling.kevin@mayo.edu) Anthony Karnezis (akarnezis@bccrc.ca) Janine Senz (janine.senz@bccrc.ca) Winnie Yang (wyang@bccrc.ca) Elena-Sophie Prigge (elena.prigge@med.uni-heidelberg.de) Miriam Reuschenbach (miriam.reuschenbach@med.uni-heidelberg.de) Magnus Von Knebel Doeberitz (knebel@med.uni-heidelberg.de) C Blake Gilks (blake.gilks@vch.ca) David G Huntsman (dhuntsma@bccancer.bc.ca) Jamie Bakkum-Gamez (bakkum.jamie@mayo.edu) Jessica N Mcalpine (jessica.mcalpine@vch.ca) Michael S Anglesio (manglesio@bccrc.ca) Version:3Date:13 April 2015 Author's response to reviews: see over
Dr. Dafne Solera Executive Editor, BMC Cancer BioMed Central Floor 6, 236 Gray's Inn Road London, WC1X 8HB United Kingdom Monday, April 13, 2015 RE: reviewer comments on manuscript # 1612983015157460 Dear Dr Solera: I am please to have received some constructive comments on our manuscript Targeted Deep Sequencing of Mucinous Ovarian Tumors Reveals multiple overlapping RAS-Pathway activating mutations in Borderline and Cancerous Neoplasms. I would like to thank reviewer 1 for providing valuable feedback and constructive comments. Please find below our response to comments and critics from both reviewer along with detailed point-by-point modifications to the manuscript, where applicable. An updated manuscript has been uploaded to your web site. I would appreciate a confirmation on the receipt of these documents. Sincerely, Michael Anglesio, PhD University of British Columbia Department of Pathology and Laboratory Medicine Robert HN Ho Research Centre RH275-2660 Oak Street Vancouver, BC V6H 3Z6
Below are the comments from reviewers along with our response and point-by-point list of any changes to the manuscript. Changes in the manuscript and accompanying documents are highlighted through the use of RED coloured text where applicable. Response to reviewer 1: MINOR ESSENTIAL REVISIONS 1. Abstract should clearly state primary and secondary aims of work. The abstract has been modified to state the hypothesis of the current study more distinctly: Our goals were to establish the overall frequency of cancer-hotspot mutations across a large cohort, especially those tumours previously thought to be RAS-pathway alteration negative, using highly sensitive next-generation sequencing as well as further explore a small number of cases with apparent heterogeneity in RAS-pathway activating alterations. Additional small changes to abstract have been made to improve clarity DISCRETIONARY REVISIONS 2. The MS is data-dense which may inhibit reading by the general reader. The MS might be more accessible if you had a summary diagram for Figure 1 showing the experimental design (number of cases included and excluded because of sequencing failure [as recommended by REMARK guidelines] and the hierarchical classification by KRAS/ERRB2 status). We have added an outline diagram now numbered as Figure 1 to illustrate the sequencing strategy, resultant data, and define where samples were excluded based on criteria noted in the main text and methods. Other figures have now been re-numbered to account for this new figure. New Figure 1 added, a callout to this figure is noted on Page 4, line 145. Callouts to other figures have been re-numbered to reflect the additional figure. 3. We have also found it quite helpful to generate Oncoprint and MutationMapper figures from our data as these are widely understood and visually attractive. (http://www.cbioportal.org/oncoprinter.jsp http://www.cbioportal.org/mutation_mapper.jsp) In order to maintain the presentation of data grouped into previously defined subcategories we have not generated oncoprint figures, however we would like to indicate that the display of data in figures 2 and 3 (previously figures 1 and 2) is similar to that of an oncoprint. We have also modified the figure legends in a way which we feel will make the data more comprehensible. Modified figure legends to Figures 2 and 3
4. You might want to speculate in the discussion about what genetic or environmental events are needed for MBT to evolve into MC. As in Barrett's oesophagus and ovarian p53 signature lesions, the key mutations are already present in some MBT, but there is no invasive disease. How is this explained? We suspect that acquisition of p53 mutations is a later event, similarly secondary RAS alterations are likely to contribute to the transition of borderline to invasive disease. Despite this it seems clear that mucinous ovarian cancer belong to a class of diseases wherein somatic activation of the RAS pathway is an obligate characteristic Somatic RASopathy. Unfortunately due to the relative rarity of this tumour and the recent modernization of histotype differentiation incorporating molecular data no substantial epidemiological studies have been undertaken that use modern diagnostic criteria, it is therefore difficult to comment on environmental, behavioral, or cultural influence. Nonetheless, given what is known about RAS-activation and cellular senescence it is plausible that p53 mutation may impart genomic instability that in turn leads to accumulation of other mutations permissive overcoming senescence and other anti-growth signals regulated by constitutive RAS-activation (for example acquisition of PTEN loss of function mutations). Unfortunately a larger genome-wide study was not possible in this cohort though it would be interesting to evaluate accumulated DNA copy number changes and clonality between MBOT and MUC especially correlating genomic complexity with acquisition of secondary RAS-mutations, and p53 mutation acquisition. Keeping in mind such a study would require a relatively large cohort, certainly larger than the one available here. We have modified the last paragraph of the discussion, Page 12, Line 320 to reflect the comment above. Response to reviewer 2: 1. The authors should clearly state major conclusion by raising results. Major conclusion is ambiguous and these data is of minor benefit for patients. We feel the changes requested by reviewer 1 should be sufficient to clarify the hypothesis and conclusions of the manuscript. Further we do not agree that this is of limited benefit, though no direct target is or therapeutic strategy is highlighted we have made a significant finding around the presence of apparently non-redundant RAS-pathway alterations. It should be clear to clinicians undertaking a strategy of targeted therapeutics that either clear or cryptic alternative mechanism of RAS-activation may be present and single agent targeted therapeutic strategies should be considered with caution in the case of aggressive or recurrent disease. We have made addition to the text in order to clarify the above conclusions. Page 4, line 132.
2. Please show the evidence that concurrent ERBB2 amplification and KRAS mutation play a critical role for tumor progression in mucinous ovarian tumors. For example, the authors should demonstrate whether concurrent ERBB2 amplification and KRAS mutation correlate with aspects of mucinous ovarian cancer survival using Kaplan Meier analysis. There is no claim that specifically ERBB2 amplification and/or KRAS mutation play critical roles in tumour progression. In contrast, the frequency of RAS-activating alterations, as a whole (including amplification of ERBB2 and mutation of KRAS, amongst others), is clearly an essential part of the biology of this tumour type. With only a single case of MUC or MBOT not found to be affected by a RAS activating alteration it would seem foolish to deny this, it is not only critical but apparently obligate. No survival difference is seen between any one RAS-alteration and any other, nor between single mutant and double mutant case this is stated in text. However one cannot conclude that two RAS activating mutation do not have more detrimental effect than single because the study is not adequately powered to address this question. To present Kaplan-Meier analysis on any of the above noted data would be irresponsible and at best misleading. 3. The discussion and conclusion are too long. This would appear to be in contradiction to the previous reviewer s comment on the manuscript being data-dense. Nonetheless we do not feel there is redundancy in the text and no changes have been made. 4. Further to confidential comments to the editor: Although the manuscript appears to be totally well-prepared, the authors essentially displayed the enumeration of gene expression data. Furthermore, most of the data in this paper is repetitive of the already published data, and the key evidence for supporting the major conclusion could have been predicted from the literature. This manuscript provides limited data, suggesting this manuscript to be unsuitable for publishing as article containing new and useful findings in this journal. This above confidential comments appears to have been forwarded in error, nonetheless we strongly disagree with the statements above, in particular: It is unclear what is meant by the reviewer on their comments essentially displayed the enumeration of gene expression data as we have not discussed gene expression data, nor presented any in the manuscript. This appears to be a significant misunderstanding of the reviewer s interpretation of the study data. We have confirmed relative frequencies of some known major RAS pathway effectors mutations (KRAS) and or amplifications (ERBB2). However, reports of RAS effector alterations outside of KRAS and ERBB2 in mucinous carcinoma are few and far between and our study is the largest of it s kind in applying a uniform analysis strategy.
Our finding of RAS-alteration double-hit cases is not entirely novel, having reported this ourselves previously. However defining a subset of tumour cells carrying two such alterations simultaneously is entirely novel and counter-intuitive given current literature on RAS-altered tumours (including, but not exclusive to mucinous ovarian carcinoma). This study provides definitive evidence that, similar to in vitro systems, RAS-alterations in vivo are not functionally equivalent. Functional equivalence amongst RAS-alterations has previously been a defacto conclusion of studies showing mutual exclusivity between RAS-alterations observed in tumour samples. Critically we have illustrated that (1) RAS-alterations are near ubiquitous in this family of tumours, (2) RAS-alteration are not functionally equivalent in vivo, and (3) acquisition of p53 mutations and secondary RAS-alterations is correlated with invasive behaviour in mucinous tumours. Additional Changes to the text: 1. A small number of typographical errors have been corrected throughout the text, as noted above changes are marked in red coloured text.