Cancer is associated with inferior outcome in patients with ischemic stroke

Figures

• 

  Fig. 1. Consort chart. Shown is the selection process for identification of the study population

• 

  Fig. 2. Characteristics of patients with known cancer diagnosed up to 5 years prior to stroke. The pie chart shows proportions of cancer types as specified by the respective colors and data outlined in column 1 of the table. Included were patients diagnosed up to 5 years prior to ischemic stroke as well as during the period of hospitalization for stroke. For the 12 patients with cancer diagnosed during hospitalization, the respective tumor types are listed in column 3. Column 4 shows how many patients received any diagnosis of any second cancer at any time point prior to stroke. The last column indicates how many patients of the respective cancer type died during follow-up

• 

  Fig. 3. Stroke etiology and imaging parameters of patients with ischemic stroke without and with known cancer. a–c. Shown are stroke etiology as classified by TOAST (a), incidence of ischemic lesions in less than 2 or 2 or more vessel territories (b) and of large vessel occlusions (c) for patients (%) without (grey bars) versus with (black bars) cancer diagnosed up to 5 years prior to stroke

• 

  Fig. 4. Laboratory parameters of patients with ischemic stroke without and with known cancer. a–f Laboratory parameters (first measurement available after admission) of 753 patients with ischemic stroke; 694 patients without (grey symbols/bars) and 59 patients with known cancer (black symbols/bars) were analyzed for an association with cancer status. Shown are data including means and SD analyzed by Mann–Whitney U test (a, c–f) or percentages of categorial variables analysed by Chi-square test (b) for the number of patients without and with cancer who had data on the indicated laboratory parameters available based on retrospective chart review: hemoglobin (a, n = 694 and n = 59 patients), white blood count and platelet count (b, n = 694 and n = 59 patients). D-dimers (c, n = 487 and n = 38 patients), erythrocyte sedimentation rate (d, n = 463 and n = 40 patients), C-reactive protein (e, n = 690 and n = 59 patients), lactate dehydrogenase (f, n = 574 and n = 52 patients)

• 

  Fig. 5. Clinical course and outcome of patients with ischemic stroke without and with known cancer. a–g Shown are modified Rankin scales (mRS) distributions (%) prior to stroke (a), on admission for stroke (b), and at follow-up after stroke (c) in patients without (grey bars) versus with (black bars) cancer diagnosed up to 5 years prior to stroke. Data for follow-up were included if available within 60 to 120 days after stroke or if the patient died within 120 days after stroke. Follow-up information on these patients were available by clinical visit for 71% and 50%, by indirect information via phone, relatives or documents for 20% and 23% and by documented death during hospitalization for 9% and 27% of patients without and with known cancer, respectively. NIHSS levels and their median in patients without known cancer (grey symbols, data available for n = 680 on admission, n = 624 after ~ 24 h) or known cancer (black symbols, data available for n = 59 on admission, n = 50 after ~ 24 h) (d). Kaplan–Meier survival curves of patients without (continuous line) or with (dashed line) cancer for the entire cohort (e) and separately for men (f) and women (g)