Predictors of Outcome in Patients with Medullary Hemorrhage.

BACKGROUND

Isolated medullary hemorrhage (MH) is an uncommon presentation of spontaneous intraparenchymal hemorrhage. The relationship between MH and neurological outcome is not well known. This study aims to assess predictive parameters for the outcome of medullary hemorrhage.

METHODS

We conducted an extensive search of the literature for cases with spontaneous, isolated MH. The study was conducted according to the statement of Preferred Reporting Items for Systematic Reviews and Meta Analysis (PRISMA). Forty-three cases diagnosed by CT or MRI have been reported in the literature, to which we add three confirmed by MRI. The ventrodorsal size of hemorrhage was taken into account as a parameter of outcome. Early neurologic deterioration (END) was defined as an incremental increase in the National Institutes of Health Stroke Scale score by ≥1 point in motor power, or ≥2 points in the total score within the first week after admission. Modified Rankin Score (mRS) 0-2 was presumed as favorable outcome (FO) and mRS 3-6 score as unfavorable outcome (UO) at discharge and at 3 months after stroke.

RESULTS

We enrolled 46 patients, and 17 (37%) patients were diagnosed with END. The cause of medullary hemorrhage was mostly vascular malformations, including cavernous malformation (33%) and arteriovenous malformation (11%). In univariate analyses, neither arteriovenous malformations (OR, 1.04; 95%CI, 0.10-10.53; P = 0.68) nor cavernomas (OR, 1.04; 95%CI, 0.22-4.89; P = 0.62) were associated with UO. Acute respiratory distress syndrome was higher in patients with UO group (44%) compared to those with FO group (16%), but this difference did not reach to a significant level (OR, 4.13; 95%CI, 0.85-20.04; P = 0.09). The diameter of hemorrhage was significantly larger (≥1 cm) in patients with UO compared to those with FO (OR, 16.67; 95%CI, 1.87-148.89; P = 0.003). Three months after stroke, 37 patients (80%) had FO and 9 (19.5%) had UO, and 5 (11%) died. Multiple logistic regression analysis using predetermined variables found to be significant in univariate analyses (END, consciousness disturbance at admission, hemorrhage size, and hypertension) showed that END at stroke onset was significantly associated with UO (OR, 4.97; CI95%, 1.13-21.94; P = 0.03).

CONCLUSIONS

These results suggest that the END is a predictor for UO in patients with medullary hemorrhage. The extent of the medullary hemorrhage along the conduction tract may contribute to deterioration.