To the Editors:

We agree with Peparini and Chirletti’s remark on our article that tumor satellites are an important issue in the field of colorectal cancer.1 There is growing evidence that these deposits are of high prognostic relevance. However, the current 7th edition of the tumor, node, metastasis system (TNM) classification was not yet available when we submitted our manuscript for consideration, and therefore it is not applicable to our study.2

It is correct that sentinel lymph node (SLN) mapping does not facilitate the detection of deposits. Moreover, when the lymph node examination is limited to the SLN, tumor satellites would probably go undetected. Our concept of ex vivo SLN mapping, however, does not influence standard histopathological examination, which includes tumor sampling and the evaluation of all nodular structures.1

The classification of tumor deposits was changed in each of the last three TNM editions and remains controversial. Therefore, the suitability of these classifications for scientific purposes seems questionable. We investigated the origin of tumor deposits in a recently published study.3 In 50% of the deposits, a distinct origin (continuous growth, venous, lymphatic, nerve sheath invasion) could be found in corresponding step sections. Neither size nor shape of the satellites was associated with their origin. In other words, in none of the last three TNM editions were the majority of deposits correctly classified. The changes between the editions cause dramatic stage shifts, reducing the comparability of cases classified according to different editions.

Interestingly, Puppa et al. hypothesized that at least a part of tumor deposits represent a special type of aggressive vascular invasion with similarity to in-transit metastasis of malignant melanoma. Therefore, they propose a classification as pM1a.4

In conclusion, the problem of correct tumor deposit classification is not sufficiently solved yet.