Clinical and Neuroimaging Features of Acute Ischemic Stroke in Cancer Patients

Background: The occurrence of acute ischemic stroke in cancer patients is not unusual. In clinical practice, acute ischemic stroke with cancer usually cannot be diagnosed promptly due to lack of specific markers. But for cancer patients, advanced prevention, accurate diagnosis and proper treatment of acute ischemic stroke are very important. The aim of the present study was to investigate the clinical and neuroimaging features of acute ischemic stroke in patients with cancer. Methods: We conducted a retrospective review of all cancer-associated acute ischemic stroke patients (n = 46) admitted to the Affiliated Hospital of Academy of Military Medical Sciences between October 2011 and March 2015. A group of non-cancer acute ischemic stroke patients (n = 50) at the same period were selected randomly as control. The clinical and neuroimaging data were collected and compared between the 2 groups. Results: Patients with cancer-associated stroke (CS) had a lower body mass index (23.26 +/- 3.70 vs. 24.88 +/- 2.83, p = 0.021) compared to non-cancer stroke (NC) patients. A lower proportion of CS patients suffered from hypertension (45.7 vs. 68.0%, p = 0.039) and hyperlipidemia (10.9 vs. 72.0%, p = 0.000) than the NC group. A higher proportion of CS patients had deep vein catheter (24.0 vs. 0%) before the onset of stoke than that of the NC group. Levels of hemoglobin, albumin and triglyceride were lower in CS groups compared with that of the NC group (p < 0.05). The prothrombin time, international normalized ratio, D- dimer and fibrinogen levels were significantly higher in the CS group than in the NC group (p < 0.05). As to the neuroimaging patterns, disperse lesions (OR 7.01; 95% CI 1.17-42.12; p < 0.05) was independently associated with CS. Conclusions: Cancer-associated ischemic stroke was different form conventional ischemic stroke in the aspect of clinical and neuroimaging manifestation. This phenomenon might be because of the embolic etiology of CS. These features together could become a clue to CS. (C) 2016 S. Karger AG, Basel