To the editor:

We thank Yamada et al. for their constructive comments entitled “Is serum HER2 ECD a predictive biomarker for response to trastuzumab in advanced gastric cancer?” related to our article [1].

In our study, an objective response rate (ORR) was not shown by subgroups because most patients (48 of 65) were treated with first-line chemotherapy. In those 48 patients, those with higher baseline serum human epidermal growth factor receptor 2 (HER2) extracellular domain (ECD) levels (15.0 ng/mL or more and 10.65 ng/mL or more) had a better ORR (89.5 % vs 52.0 %, P = 0.01; 81.3 % vs 33.3 %, P = 0.004) than those with normal baseline serum HER2 ECD levels (less than 15.0 ng/mL and less than 10.65 ng/mL). Because of the limited number of patients, in the 14 patients treated with second-line trastuzumab-based chemotherapy, a higher baseline serum HER2 ECD level was still associated with better ORR although not to a statistically significant extent.

We found a high baseline serum HER2 ECD level and a decline of serum HER2 ECD level during chemotherapy significantly correlated with the response to chemotherapy, which is consistent with the findings reported in another article concerning gastric cancer [2]. In breast cancer, there are indeed contrary results concerning the predictive value of serum HER2 ECD levels in HER2-positive breast cancer [3]. However, the mechanism is still unclear.

In our observational study we prospectively collected consecutive HER2-positive advanced gastric cancer patients treated with trastuzumab-based chemotherapy. We could not exclude the potential bias due to the heterogeneity of the patients. As we discussed in the “Discussion” section, further well-designed prospective studies with a larger sample size need to be performed to validate and strengthen our conclusion.