Carotid MR Imaging Combined With Laboratory Tests And ESRS Score In Prediction Study For Recurrent Acute Ischemic Stroke

Objective: To develop an optimized stroke risk score system– revised stroke risk score(RSRS), based on carotid plaque high-resolution magnetic resonance imaging(MRI) combined with laboratory tests and Essen Stroke Risk Score(ESRS), and analyze the predictive value of RSRS in prediction for recurrent acute ischemic stroke(AIS) by compared with ESRS.Materials and methods: Forty-one patients with AIS proved by clinic and follow-up, including initial AIS(n=20) and recurrent AIS(n=21), were enrolled in this retrospective and ethics committee-approved study. All patients’ clinical information, laboratory tests and radiological information were collected and analyzed. Carotid plaque high-resolution MR imaging was performed in all patients, and the responsible carotid arteries were quantitatively and qualitatively analyzed. The carotid plaque burden and compositional features were compared between the initial and recurrent stroke patients. All the potential risk components in relation to recurrent AIS were determined by multiple Logistic aggression analysis. The best weighting scheme for all independent predictors were determined and enrolled into RSRS. The predictive value of RSRS in prediction for recurrent AIS was analyzed by compared with ESRS using receiver operating characteristic curve(ROC).Results:(1) The prevalence of hypertension and diabetes mellitus in patients with recurrent AIS were significantly greater than patients with initial AIS(P = 0.023 and 0.003, respectively, both P < 0.05), whereas other potential components including age, gender, heart disease, smoking and peripheral arterial disease showing statistically insignificant between the initial and recurrent AIS patients(all P > 0.05).(2) The level of low-density lipoprotein cholesterol(LDL-C) and total cholesterol(TC) in patients with recurrent AIS were significantly greater than patients with initial AIS(P = 0.020 and 0.004, respectively, both P < 0.05), and there existed no statistically significant between initial and recurrent AIS patients in other laboratory tests including fibrinogen(FIB), homocysteine(HCY), high-density lipoprotein cholesterol(HDL-C) and triglyceride(TG)(all P > 0.05).(3) The mean NWI were greater in patients with recurrent AIS than those with initial AIS(P =0.038, P < 0.05). However, we found no statistical significance in lumen area(LA), wall area(WA) and total vessel area(TVA) between the two groups(all P > 0.05). The mean lumen stenosis in patients with recurrent AIS was higher, but showed insignificant statistically than that of patients with initial AIS(P > 0.05).(4) Patients with recurrent AIS had a significantly higher prevalence of fibrous cap rupture(FCR) and intra-plaque hemorrhage(IPH) compared with those with initial AIS(P =0.018 and 0.029, respectively, both P < 0.05). And the prevalence of lipid rich necrotic core(LRNC), calcification and lumen stenosis >50% in patients with recurrent AIS showed statistically insignificant compared with those with initial AIS(all P > 0.05). The LRNC index in recurrent AIS patients was statistically higher than that in initial AIS patients(P = 0.030,P < 0.05), but we found statistical insignificance in the IPH and calcification index between two groups(all P > 0.05).(5) LDL-C, IPH, FCR, NWI and LRNC index were independent risk factors in prediction of recurrent AIS.(6) There were statistically significant in both ESRS and RSRS between initial and recurrent patients(both P < 0.001). And there also existed statistically significant in both initial and recurrent AIS patients between two score systems(both P < 0.001).(7) ROC of both ESRS and RSRS were made, AUC of ESRS was 0.890(95%CI, 0.794-0.987), but AUC of RSRS was 0.912(95%CI, 0.820-1.000). And ROC of RSRS located in the upper-left of primary ESRS, indicating RSRS was better than ESRS on the recurrence risk evaluation.(8) ESRS = 3 scores or RSRS = 6 scores were the critical value for prediction of high-risk population of recurrence. Conclusion:(1) LDL-C, IPH, FCR, NWI and large area of LRNC were independent risk factors in prediction of recurrent AIS.(2) There existed bigger plaque burden and more vulnerable compositions in recurrent AIS patients.(3) Both ESRS and RSRS were valuable in prediction for recurrent AIS patients, but ESRS was much more valuable than ESRS.(4) ESRS = 3 scores or RSRS = 6 scores were the critical value for prediction of high-risk population of recurrence.