Clinical and imaging characteristics of cryptogenic stroke combined with right-to-left shunt

Jinshan YANG; Yifan LIN; Lihua CHEN; Huixing WEI

ABSTRACT

Objective:

To investigate the clinical and imaging characteristics of patients with cryptogenic stroke with different right-to-left shunt (RLS) and possible mechanisms.

Methods:

Acute ischemic stroke patients aged 18-65 years, who admitted to the Department of Neurology, the First Affiliated Hospital of Fujian Medical University from April 2012 to July 2022, were etiologically classified according to the TOAST criteria. Cryptogenic stroke patients further underwent contrast-enhanced transcranial Doppler (cTCD) examination. Patients with RLS (RLS grade Ⅰ-Ⅳ) were divided into large shunt group (RLS grade Ⅲ-Ⅳ, under normal breathing or after Valsalva manoeuvre) and small shunt group (RLS grade Ⅰ-Ⅱ, under normal breathing or after Valsalva manoeuvre). The chi-square test,Fisher's exact probability method and one-way analysis of variance were used to compare clinical and neuroimaging data between the 2 groups. And the RLS related factors as well as mechanisms of RLS-related stroke were further discussed through Spearman correlation and multiple Logistic regression analysis.

Results:

A total of 167 cryptogenic stroke patients with RLS were included in this study. The male to female ratio was 88∶79. The onest age was (45.27±9.02) years. In the large shunt group, there were 92 patients (55.1%) with onset age of (45.61±9.32) years. In this group, 35 patients (38.0%) were males, 9 (9.8%) had hypertension, and 18 (19.6%) had smoking history; the Risk of Paradoxical Embolism (RoPE) score was 7.08±0.12. In the small shunt group, there were 75 patients (44.9%) with onest age of (44.85±8.68) years. In this group, 53 patients (70.7%) were males, 29 (38.7%) had hypertension, and 31 (41.3%) had smoking history; the RoPE score was 6.20±0.14. Males (χ 2=17.640, P<0.001) as well as hypertension (χ 2=19.611, P<0.001) and smoking history (χ 2=9.443, P<0.01) were significantly less common in the large shunt group than in the small shunt group, while the RoPE score in the large shunt group was higher than that in the small shunt group ( F=21.580, P<0.001). All 6 pulmonary arteriovenous fistula cases (6.5%, 6/92) belonged to the large shunt group. There was no pulmonary arteriovenous fistula patient in the small shunt group ( P<0.05). There were no statistically significant differences in age of onset, hyperlipidemia, diabetes, previous stroke/transient ischemic attack, or family history of stroke between the 2 groups (all P>0.05). Multiple Logistic regression analysis showed that female ( OR=2.848, 95 %CI 1.223-6.632, P<0.05) and no history of hypertension ( OR=3.937, 95 %CI 1.593-9.727, P<0.01) were significantly associated with large shunt. The neuroimaging features were analyzed through comparing offending vessels as well as spatial distribution of lesions between the 2 groups.

Results:

showed that infarct location of 54 cases (58.7%) was in the unilateral anterior circulation area, 34 cases (37.0%) in the posterior circulation area, no case in the bilateral anterior circulation area, and both of anterior and posterior circulation areas were involved in 4 cases (4.3%) in the large shunt group. Of all the large shunt group patients, 31 (33.7%) had cortical infarcts, 9 (9.8%) had subcortical infarcts, 36 (39.1%) had deep infarcts, and 16 (17.4%) had mixed multiple infarcts. In the small shunt group, infarct location of 36 cases (48.0%) was in the unilateral anterior circulation area, 29 cases (38.7%) in the posterior circulation area, 4 cases (5.3%) in the bilateral anterior circulation area, and both of anterior and posterior circulation areas were involved in 7 cases (9.3%). Of all the small shunt group patients, 7 (9.3%) had cortical infarcts, 8 (10.7%) had subcortical infarcts, 39 (52.0%) had deep infarcts, and 21 (28.0%) had mixed multiple infarcts. Logistic regression analysis showed that cortical infarction was associated with large shunt ( OR=5.981,95 %CI 1.882-19.007, P<0.01), and there was no statistically significant difference in other types of infarction distribution between the 2 groups (all P>0.05).

Conclusions:

Among cryptogenic stroke patients with RLS, large shunt was more common in females, patients with no history of hypertension, and those with higher RoPE score as well as more cortical infarction, suggesting that there was a strong link between stroke and large RLS, which revealed the underlying mechanism of embolism in patients with stroke and large RLS. In the small shunt group, there were more males and hypertensive patients, suggesting atherosclerosis was more likely involved in pathogenic mechanism of stroke patients with small RLS.