We appreciate the letter by Chandiramani and Kenny bringing our attention to 2 areas of concern in our recent article that would benefit from our clarification.
Regarding the first point, we would like to reassure our readers that our sequencing strategy, using both polymerase chain reaction and Sanger sequencing, was sufficient to detect any existing mutations in the exome domains, and if mutations at codon 308 or 409fs loci existed in our cohort of Chinese patients, we should have been able to detect them. There is another line of evidence from a study by Gonzalez-Bosquet et al in which 2 cases of a GATA3 (GATA-binding protein 3) point mutation were observed in the 60 breast cancer samples that underwent bidirectional sequence analysis; no cases of gene insertion or deletion were found in the GATA3 exome domain. Furthermore, we have conducted GATA3 mutation analysis in an additional validation cohort of 153 patient samples, and observed a similar mutation profile. In this validation cohort, somatic mutations in GATA3 were detected in 11.8% of tumors (18 of 153 tumors), 77.8% of which (14 tumors) were missense point mutations whereas 11.1% (2 tumors) were frameshift insertions and 11.1% (2 tumors) were frameshift deletions. We believe that the fundamental biological differences between breast cancers occurring in Chinese compared with Western populations best explain the difference in GATA3 mutation profiles. We would like to emphasize that the “Asian” category in The Cancer Genome Atlas (TCGA) database includes a wide range of ethnic groups, and the percentage of “Chinese” patients included is unknown. This may account for the difference between TCGA and the findings of our study.
Although we stated in our original article that the follow-up period of patients in the TCGA cohort ranged from 1 month to 226.5 months, a clarification we did not make clearly was that in our survival analysis of the TCGA cohort, we included only those patients whose follow-up data exceeded 1 month, excluding all patients who lacked follow-up data. It is interesting to note that our conclusion regarding improved survival in patients with luminal A breast cancer with a GATA3 mutation was mainly drawn from data of the Fudan University Shanghai Cancer Center cohort, because the TCGA database lacked comprehensive follow-up data and only provided statistics regarding overall survival. In addition, the TCGA database is continuously being updated, and newly included data may cause current analysis results to differ from previously conducted analysis.
We therefore maintain our main conclusion that GATA3 mutations in patients with luminal-like breast cancer may define a subgroup, and fully acknowledge that further studies are needed to validate these findings.