To the Editor,

Many thanks to Dr Altundang et al. for your comments about reproducibility of Ki67 usefulness in clinical practice.

It is true that the American Comprehensive Cancer Network Guidelines do not provide any information regarding Ki67 immunohistochemical assessment. However, the Saint Gallen Consesus meeting [1,2,3] had suggested several times the use of Ki67 for classifying Luminal breast carcinomas since 2011. Different cut-off values have been proposed: initially it was 14% and more recently it was upgraded to 20%, even Ki67 as a continuous variable has been suggested. Cheang et al. [4] used the Ki67 cut-off value of 14% to discriminate Luminal A and B molecular subtypes. Our study confirms that a cut-off value of 14% and Ki67 as a continuous variable can predict the prognosis of patients with luminal breast carcinoma. Furthermore, the use of this marker of proliferation is very common in breast and many other areas of pathology and it is frequently asked for oncologist before deciding several treatments, despite the well-known problems in their reproducibility, related with antibodies, methodological differences in inmmunohistochemistry, and the pathological evaluation.

Finally, we agree with you that reproducibility is a major concern about Ki67, but it was not the subject of our study. Of note, we expect that future recommendations based on digital pathology and better standardization of the technic will allow more accurate results and reproducibility in clinical practice.