Prediction of deep vein thrombosis by ultrasonography and D-dimer in Asian patients with ischemic stroke

Ischemic stroke survivors demonstrate a high rate of neurological deficit, with a considerable proportion of patients remaining immobilized for a period after stroke. Immobility after ischemic stroke can cause various complications, such as deep vein thrombosis (DVT), pressure ulcers, or sarcopenia leading to falls and fractures [1]. DVT and pulmonary embolism are leading causes of death after stroke [2] and are the most common causes of death during the period of active rehabilitation, [3] with newly-developed DVT increasing the rate of mortality at 3 months after stroke [4].

The incidence of DVT differs according to the modality used to detect its presence after stroke, [5] with previous studies reporting prevalence rates ranging from 20 to 70%. Several studies have shown that the incidence of DVT may differ according to ethnicity and race, with the incidence of DVT being lower than expected in studies on Asian patients with stroke (4.8 to 45%) [6]. The prevalence of DVT in the general population is also lower in Asian populations than in non-Asian patients [7]. However, the annual incidence of DVT in the general population is increasing in Asia, [8] and DVT remains an important issue for determining the treatment strategy for stroke, especially regarding the use of antithrombotics.

Although the D-dimer level may be useful for detecting DVT in various conditions, the initial D-dimer level has shown limitations for predicting the presence of DVT among acute ischemic stroke patients [9]. In the present study, we used lower extremity ultrasonography (LEUS) to evaluate the prevalence of DVT after ischemic stroke in Korean patients and investigated the factors associated with the presence of DVT to reveal the optimal strategy for identifying DVT in Asian stroke patients.

During the study period, 326 patients in the acute phase within 7 days from stroke onset were admitted to the stroke center of Kyung Hee University Hospital. Of these patients, 37 were excluded because of poor neuroimaging quality (n = 5) or the absence of LEUS data (n = 32). Finally, 289 patients were analyzed. The D-dimer level was checked in 283 patients. The mean age of the patients enrolled in the study was 68.1 ± 11.3 years, and 175 (60.6%) of the patients were male. None of the patients were pregnant or under contraceptive or other hormone replacement treatment. One patient was diagnosed with anti-phospholipid syndrome, but did not show a DVT. Two patients showed mRS 4 prior to the index stroke with moderate disability, which needs assistant to walk, but did not show DVT.

In this study, LEUS revealed a 13% incidence of DVT among patients with ischemic stroke. Female sex and a severe neurological deficit were independently associated with the presence of DVT in the acute phase of ischemic stroke. However, the usefulness of D-dimer level for screening DVT in Asian patients with acute ischemic stroke was limited. Elevated D-dimer level was observed in more than a quarter of patients with ischemic stroke, with the mechanism of stroke being particularly associated with elevated D-dimer level.

Although D-dimer level is useful for detecting DVT in other conditions such as liver disease and malignancy, several studies showed that the initial D-dimer level is not associated with the presence of DVT in stroke patients [9]. Ischemic stroke shows associations with thrombotic/fibrinolytic conditions, especially in the case of cardioembolic and cancer-associated strokes [13]. Therefore, the D-dimer level is typically increased in those with cardioembolic strokes [14] and is useful for detecting cancer-associated strokes [15]. In accord with previous studies, our results also showed that D-dimer level was associated with the stroke mechanism, with cardioembolic stroke being an independent factor associated with elevated D-dimer levels. As D-dimer is elevated in acute ischemic stroke patients and as thrombolysis and the stroke mechanism influences the D-dimer level, it may be less useful for detecting DVT in patients with acute ischemic stroke.

Instead, severe neurological deficits were associated with DVT in acute ischemic stroke patients. There was a significant difference in the NIHSS score between those with and without DVT. Especially the NIHSS score of the lower leg weakness, which directly influences the ambulation in their acute period, showed a significant difference. It is well known that immobilization or paralysis of the lower extremities is associated with DVT because of the increased venous stasis [16]. The prevalence of DVT or pulmonary embolism is lower in Asians than in non-Asian subjects, and coagulopathy is less associated with ischemic stroke in Asian patients [17, 18]. Therefore, the relative importance of systemic coagulopathy represented by the D-dimer level may be less important in Asian patients with ischemic stroke, and immobilization and severe neurological deficit may be more important in the development of DVT in Asian patients with acute ischemic stroke [19].

The results of our study indicate that the role of D-dimer for the screening of DVT in Asian patients with acute ischemic stroke may be limited. Studies using D-dimer to screen for DVT showed a very low rate of DVT in Asian patients with acute ischemic stroke [20]. Therefore, the use of LEUS may be more appropriate in selected high-risk patients. LEUS for the detection of DVT also has the benefits of no radiation hazard and cost-effectiveness in comparison with computerized tomography venography. High-risk patients are those with a severe neurological deficit or immobilization, and acute ischemic stroke patients with a low D-dimer level should not be presumed to be free of DVT [21]; in our study, more than 60% of patients with DVT showed a normal D-dimer level. Furthermore, performing LEUS on these patients provided additional information on the size, chronicity, and degree of occlusion of the thrombus, which may help in determining the management of DVT in patients with acute ischemic stroke. However, LEUS is observer dependent and still has limitations in detecting DVT in the pelvic or calf area in comparison to computerized tomography venography [12]. Therefore, the evaluation for detecting DVT in acute ischemic stroke patients may be individualized. Furthermore, in those with high risk DVT, prophylaxis with intermittent pneumatic compression, unfractionated heparin or low-molecular weight heparin must be also individualized based on the risk of intracerebral hemorrhage [22].

Our study has several limitations, including those stemming from the small sample size from a single center. The LEUS was performed within 3 days from admission, which is a very early period, and the prevalence of DVT may have increased if the LEUS was performed later on after the index stroke, and the relative power of the predictors could also have been different. Furthermore, we only checked D-dimer on the day after admission, and did not make any follow-up checks. The usefulness of D-dimer at follow-up was observed in a former study [9]. Third, patients with distal DVT, for which the clinical implication is still controversial, were included in this study. Finally, only a limited number of patients received intermittent pneumatic compression, as it was not commonly used during the study period. The results may differ under a more aggressive care to prevent DVT in those with high risk patients.

Despite these limitations, our study shows that the prevalence of DVT is considerable in Asian patients with acute ischemic stroke. High neurological severity, but not the D-dimer level, was associated with the presence of DVT. Therefore, the role of D-dimer for screening for DVT in Asian patients with acute ischemic stroke may be limited, and LEUS should be considered for those patients with severe neurological deficit or immobilization.