Circular RNA FUNDC1 improves prediction of stroke associated infection in acute ischemic stroke patients with high risk

Abstract

Identifying those patients who were at high risk of stroke associated infection (SAI) for preventive antibiotic therapy was imperative for patients' benefits, thus improving prediction of SAI was critical for all acute ischemic stroke (AIS) patients. Circular RNA FUNDC1 (circFUNDC1) has been reported to be the diagnosis and prognosis biomarker of AIS. Therefore, the present study aimed to figure out whether circFUNDC1 could be the potential predictor of SAI that could help to guide preventive treatment. In total, 68 patients were included in the study, 26 of which had infection and 42 without. Copy number of circFUNDC1 in plasma were quantified by quantitative real-time polymerase chain reaction (qPCR). Platelet spike-in experiment and correlation analysis were conducted to explore possible origins of circFUNDC1 in plasma. A significantly elevated level of circFUNDC1 was found in SAI patients compared with not infected AIS patients (P=0.0258). Receiver operating characteristic (ROC) curves demonstrated the prediction significance of circFUNDC1, with the area under the curve (AUC) at 0.6612 and sensitivity, specificity at 69.23%, 61.90% respectively in predicting SAI. Then, when adding circFUNDC1 in the risk model, the AUC increased from 0.7971 in model A to 0.8038 in model B. Additionally, positive correlation was observed between circFUNDC1 level and neutrophils counts. WBC and neutrophil ratios were significantly elevated in SAI patients compared with non-SAI patients. Therefore, circFUNDC1 could be used to construct a risk model for the prediction of SAI that is beneficial for AIS patients' preventive treatment.

Keywords: acute ischemic stroke; biomarker; circular RNA; risk models; stroke associated infection.

Figure 1. Expression of circFUNDC1, cricPDS5B and circCDC14A in acute ischemic stroke (AIS) patients and predictive value of circFUNDC1 for SAI

(A–C) circFUNDC1, circPDS5B and circCDC14A levels in SAI patients and non-SAI patients, n = 26 and 42 respectively, median ± interquartile range, Mann–Whitney test. (D) Receiver operating characteristics (ROC) curve was calculated merely depending on circFUNDC1 level in predicting SAI. (E and F) ROC curves were calculated by risk model A and risk model B.

Figure 2. Kaplan–Meier curves estimates SAI and survival for AIS patients

(A) Kaplan–Meier curves estimates SAI during the first 7 days after stroke, n=34/34, P=0.0078, Log rank test. (B), Kaplan–Meier curves estimates survival till to 90 days after stroke, n=34/34, P=0.0129, Log rank test.

Figure 3. Platelet spike-in experiment as well as the correlation between platelets counts and copy number of circFUNDC1 per microliter plasma

(A) Levels of circFUNDC1 after spiking back platelets into platelet-poor plasma (PPP) compared with platelet-rich plasma (PRP), n=8. Mean ± SEM, Friedman test. (B) Correlation between platelet counts and copy number of circFUNDC1 per microliter plasma; N=66, Pearson correlation analysis.

Figure 4. Correlations between circFUNDC1 levels and red blood cell (RBC), white blood cell (WBC), neutrophil and lymphocyte counts

(A) Correlation between circFUNDC1 levels and RBC counts. (B) Correlation between circFUNDC1 levels and WBC counts. (C) Correlation between circFUNDC1 levels and neutrophil counts. (D) Correlation between circFUNDC1 levels and lymphocyte counts; N=66, Pearson correlation analysis.