| Literature DB >> 1651557 |
M Eskelinen1, P Lipponen, S Marin, H Haapasalo, K Mäkinen, H Ahtola, J Puittinen, P Nuutinen, E Alhava.
Abstract
A retrospective clinicopathologic study was done on 111 patients with a pancreatic ductal adenocarcinoma. The mean follow-up period was 6 years. By means of interactive morphometry six nuclear morphometric features were measured in biopsy specimens from the primary tumours. Volume-corrected mitotic index (M/V index) was estimated in the same sections. Histologic grading was done in accordance with the WHO. The M/V index (p = 0.002), the nuclear area of the 10 largest nuclei (NA10) (p = 0.025), the histologic grade (p = 0.0956), the nuclear area (NA) (p = 0.038), the standard deviation of the nuclear perimeter (SDPE) (p = 0.033), and the standard deviation of the nuclear area (SDNA) (p = 0.0430) predicted survival in univariate analysis. The type of surgery performed was a significant prognosticator too (p = 0.0131). A multifactor regression analysis of survival including clinical and histologic factors identified the M/V index as the most important prognosticator (p = 0.009), followed by the type of surgery performed (p = 0.022). Other histologic factors had no independent prognostic value. Our results suggest the use of morphometric features instead of the conventional histologic grading in predicting survival of pancreatic ductal adenocarcinoma.Entities:
Mesh:
Year: 1991 PMID: 1651557 DOI: 10.3109/00365529108998570
Source DB: PubMed Journal: Scand J Gastroenterol ISSN: 0036-5521 Impact factor: 2.423