1411P - Stroke as first signal of cancer - stroke unit retrospective study

Date 28 September 2014
Event ESMO 2014
Session Poster Display session
Topics Aetiology, epidemiology, screening and prevention
Basic Scientific Principles
Presenter Mariana Rodrigues
Citation Annals of Oncology (2014) 25 (suppl_4): iv486-iv493. 10.1093/annonc/mdu353
Authors M. Rodrigues1, R. Costa2, S. Mateus2, N. Ferreira2, A. Lourenço2, F. Grenho2, L. Campos2
  • 1Oncology, Hospital São Francisco Xavier, 1495-500 - Lisbon/PT
  • 2Internal Medicine, Hospital São Francisco Xavier, 1495-500 - Lisbon/PT



Access stroke as the first manifestation of cancer, evaluate cancer as a stroke risk factor, analyse clinical manifestations and outcomes of these patients (pts) and verify the differences between pts with and without neoplasic disease (NG and CG respectively).


Retrospective analysis of clinical data from 623 patients with stroke, admitted in a Stroke Unit of a Central Hospital between 01.2011 and 09.2013. We studied 560 pts who were discharged. (49 died and 14 had missing data). Data analysis using Microsoft Excel® and SPSS®.


From the 560 pts; 512 had ischemic stroke (IS) and 48 had haemorragic. 70 pts were diagnosed with active cancer (more frequent genitourinary) closely before, during or after the stroke diagnosis. 38 pts were men in CG, 264 in the NG. The mean age was higher in the NG (72y), than in the CG (68,9 y). IS was the most common in both groups (NG: 78,5%; CG: 79.5%). There was no difference in the stroke territory between both. Cardioembolic was the most frequent etiology in both groups.Cardiovascular risk factors were more prevalent in the CG, but we found no significant statistical difference (p 0,9-1,1). At the time of discharge, the mean National Institute of Health Stroke Scale value was higher in the CG:7,11 (NG:4,29)(p = 0,008) but the modified Rankin Scale was similar. D-dimers were equivalent in both groups.


We found 12.5% pts with both diagnoses. Cancer, as a hipercoagulable state, might be a risk factor for stroke, which in turn could be a first manifestation of a neoplasic disease. The use of d-dimers as a possible predictor of cancer is not possible to confirm in this sample.


All authors have declared no conflicts of interest.