When to Screen Ischaemic Stroke Patients for Cancer

Abstract

Background and Purpose: Ischemic stroke can be the first manifestation of cancer and it is therefore important to ascertain which stroke patients should be considered for cancer-diagnostic investigations. We aimed to determine the frequency of active cancer in patients with acute ischemic stroke and to compare clinical findings in stroke patients with active cancer to ischemic stroke patients with no history of cancer. Finally, we aimed to develop a predictive and feasible score for clinical use to uncover underlying malignancy. Methods: All ischemic stroke patients admitted to the stroke unit in the Department of Neurology, Haukeland University Hospital were consecutively included in the Norwegian Stroke Research Registry (NORSTROKE). Stroke etiology was determined by the Trial of Org 10172 in Acute Stroke Treatment (TOAST) criteria. Data on cancer diagnoses was obtained from patients’ medical records and the Cancer Registry of Norway. Active cancer was defined as cancer diagnosis, metastasis of known cancer, recurrent cancer or receiving cancer treatment, all within 12 months before or after the index stroke. Based on variables independently associated with active cancer, a predictive score was developed using the area under the receiver operating characteristic (AUC-ROC) curves. Bayes’ theorem was used to calculate post-test probabilities of active cancer. Results: Of the 1,646 ischemic stroke patients included, 82 (5.0%) had active cancer. Increased D-dimer (OR = 1.1, 95% CI: 1.1–1.2, p = <0.001), lower Hb (OR = 0.6, 95% CI: 0.5–0.7, p = <0.001), smoking (OR = 2.2, 95% CI: 1.2–4.3, p = 0.02) and suffering a stroke of undetermined etiology (OR = 1.9, 95% CI: 1.1–3.3, p = 0.03) were factors independently associated with active cancer. These were included in the final predictive score which gave an AUC of 0.73 (95% CI: 0.65–0.81) in patients younger than 75 years of age. Assuming the prevalence of cancer to be 5%, the score shows that if a patient fulfills all 3 score points, the probability of active cancer is 53%. Conclusions: Active cancer was found in 5% of our ischemic stroke patients. We found that a clinical score comprising elevated D-dimer ≥3 mg/L, lower Hb ≤12.0 g/dL and previous or current smoking is feasible for predicting active cancer in ischemic stroke patients.