Prognostic Impact of DNA-Image-Cytometry in Neuroendocrine (Carcinoid) Tumours

The report of Raatz, Böcking and Hauptmann [1] represents an excellent example of an in depth retrospective study on the influence of different clinico-pathological parameters including static DNA cytome-try on tumour prognosis illustrating the power but also some limitations of a comprehensive survival analysis in surgical pathology. The authors studied in total 44 neuroendocrine (car-cinoid) tumours of different localizations including in particular gastrointestinal neuroendocrine tumors from the middlegut (6 small intestine, 1 Meckel's diver-ticulum, 1 valvula Bauhini, 10 appendix, 4 appendix plus coecum, 2 colon ascendens), foregut (5 stomach , 3 papilla Vateri, 3 pancreas) and hindgut (1 colon transv./desc., 5 rectum) and few pulmonary carcinoids (3 bronchus). The H&E sections of all tumours were reviewed, classified according to the Soga and Tazawa classification (Soga and Tazawa 1971) defining 5 different growth pattern types (A solid/nodular, B trabec-ular, C tubular/acinar, D atypical, E mixed), immuno-histochemically analyzed for NSE, Chromogranin, S-100 and peptide hormones somatostatin, serotonin, glucagons, gastrin and pancreatic polypeptide, assessed by the morphometric parameters mean nuclear area and form factor and finally DNA cytometry. In the latter analysis the parameters DNA stemlines (first and second), the 5c exceeding events (5cEE), the 5c exceeding rate (5cER), the 2c deviation index (2cDI), the DNA entropy, malignancy grade and mean content were evaluated. Finally the survival of each patient was assessed by review of the hospital files, sending questionnaires to the referring doctors and contracting the registrars' office. Thus, a plethora of information was collected for a large collective of this rare tumour type, the investigations were carefully performed and thoroughly analyzed and therefore the authors first of all need to be congratulated for their very comprehensive study. This is particularly true considering the important finding that the DNA cytometry parameters 2cDI, the 5cER, the DNA malignancy grade and entropy (each with at least p < 0.005), the pattern type of the DNA histogram as well as the form factor and an increased mean nuclear area were significantly associated with a higher mortality in the univariate analysis. The importance can not be overemphasized considering the fact that previous ploidy studies being well reviewed in the paper have shown conflicting results regarding the impact of DNA measurement on the prognosis of neu-roendocrine tumours. In addition, other clinicopatho-logical parameters like poor differentiation or globlet cell type (p < 0.00001), histological type, infiltra-tive growth, local recurrence, type of operation (cura-tive/endoscopic versus palliative), localization in the …