"The past is a different country, they do things differently there": using the SEER data-base to assess prognosis in neuroendocrine tumours.

Neuroendocrine neoplasms (NENs) are still considered to be rare neoplasms, and their epidemiology has been classically studied in large population-based cancer registries. Besides the benefits and the limitations that a cancer registry may have for all the registered cancers, the National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) Programme, has a number of drawbacks for the study of NENs. The change in management of NENs, either from diagnostic or from therapeutic points of view, the role of the Ki-67 labelling index and introduction of sensitive functional imaging, along with the misclassification of the more benign types of NENs, are the main limitations of the prognostic ability of the SEER data-base, particularly when including older data.