Abstracts

Rationale: During the 2018 Summer Urban Yellow Fever outbreak in the Sao Paulo Metropolitan Area, all severely ill patients were transferred to the same tertiary referral center (Hospital das Clinicas FMUSP). Encephalopathy, cerebral edema, and seizures were very common in this patient population. Seizures appeared to be associated with poor outcome. On January 18, we instituted primary seizure prophylaxis with antiepileptic drugs for at-risk patients (patients showing signs of encephalopathy). This study evaluated the association of seizures with disease severity and the impact of primary seizure prophylaxis with antiepileptic drugs in this patient population .   Methods: We retrospectively reviewed 89 cases of yellow fever admitted to HC-FMUSP from October 2017 to May 2018. After January 18, patients with clinically relevant encephalopathy (West Haven>=II or Glasgow Coma Scale<15) were started on prophylactic AEDs. When oral route was available, a levetiracetam loading dose of 2,000 mg was followed by a daily 1,000 mg bid maintenance dose. Intravenous lacosamide (loading dose, 200 mg, maintenance 100 mg bid) was used when oral route was unavailable. Trough serum levels were drawn 23 hours after the loading dose and daily for 2-4 days (Levetiracetam seven patients, Lacosamide, two patients).We collected clinical and liver function data, and noted presence of encephalopathy and seizure occurrence. We compared venous ammonia levels and factor V activity in patients with and without seizures. We also compared seizure occurence in the patient population before and after institution of the primary seizure prophylaxis.  Results: Mean patients’ age was 43 years (Interquartile range [IQR] 31-56), with a median SAPS3 score of 59 (IQR 47-70). Median venous ammonia on hospital admission was 62 mcmol/L (IQR 49-89). Sixty-one out of 89 patients were placed on continuous veno-venous hemofiltration, in 59/61 cases before the fourth hospital day). Median Factor V activity was 37% (IQR 19-67). Compared to patients without seizures, patients with seizures had higher ammonia levels (125mcmol/L, IQR 77-197 vs. 73mcmol/L, IQR 52-122. P<0.001) and lower factor V activity (median 32%, IQR 14-46 vs. 59%, IQR 29-104, p<0.001). Considering the whole patient population, seizure incidence was similar in the periods before (30% or 6/20 patients) and after (24.6% or 17/69 patients) institution of primary seizure prophylaxis (p=0.317). Seizure occurrence did not differ in the subset of patients with encephalopathy that did not receive prophylaxis (before Jan18) (36% or 5/14 patients) and patients who received prophylactic AEDs (32% or 17/53 patients) (p=0,797). Empirical Levetiracetam levels were within therapeutic range in 3/4 patients in 10/12 measurements (median 28.6, range 14.3-80.6). Lacosamide levels were measured on seven occasions in two patients; all levels were in the subtherapeutic range (3,1-3,7mcg/dL).   Conclusions: Seizures were common in severely ill patients with yellow fever with multi-organ dysfunction in this series and were associated with higher ammonia and lower factor V levels, indicative of liver dysfunction. The proposed primary seizure prophylaxis regimen was ineffective to prevent seizures in this patient population. Future studies should evaluate the use of prophylatic AEDs in higher doses and earlier in the disease course.  Funding: No funding