We thank Pantaleo and colleagues for their comments. They emphasize the divergent biologic nature of wild versus mutant KIT/PDGFRA GIST and believe that this key point should be addressed in more detail. We totally agree that the mutational status is important. For this reason, we used a well-defined cohort with KIT/PDGFRA mutational data available in all cases [as stated in the article (1)], which are also presented in Table 1 in our article. Our cohort has been published in several previous studies and the references are provided in the article.
In this publication, we aimed to identify candidate genes in well-known recurrent regions of copy number variations in addition to KIT and PDGFRA. The chromosomal positions of recurrent copy number regions of our study are in accordance with other publications. As stated in the article, we correlated copy number variations with clinic-pathologic data including KIT/PDGFRA mutation status. Here, we found only a correlation of 22q deletion with the presence of KIT mutations. In a next step, we looked for potential candidates within the regions of copy number variations. We evaluated the significance of candidate genes in our cohort and again correlated the findings with clinic-pathologic data including KIT/PDGFRA mutation status. So, we used statistical analysis using χ2 tests to correlate our findings with KIT/PDGFRA mutation status, and mutation status was also included into Cox regression models.
In conclusion, KIT/PDGFRA mutation data are available in all cases. We provided the data in a collective manner using the Kaplan–Meier curves, etc., and did not present the mutational status of patients evaluated in each test (as requested by Panataleo) as we believe that this approach would result in an overwhelming and confusing amount of single data. Nevertheless, protein expression of none of the genes investigated correlated with KIT/PDGFRA mutation status except immunohistochemistry for Kit, as stated in the article. So KIT and PDGFRA mutation status seem not to be of striking relevance for expression or possible GIST subgrouping in context with these genes.
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
- Received January 8, 2014.
- Accepted January 14, 2014.
- ©2014 American Association for Cancer Research.