Neuropathology of yellow fever autopsy cases.

Background: Yellow fever is a viral hemorrhagic fever caused by yellow fever virus, a mosquito-borne flavivirus. Despite an effective vaccine, major outbreaks continue to occur around the world. Even though it is not a proven neurotropic virus, neurological symptoms in more severe clinical forms are frequent. The understanding of this apparent paradox is still rarely addressed in literature.

Methods: The brains of thirty-eight patients with yellow fever confirmed by RT-PCR, who underwent autopsy, were analyzed morphologically to identify and characterize neuropathological changes. The data were compared with brains collected from individuals without the disease, as a control group. Both cases and controls were subdivided according to the presence or absence of co-concurrent septic shock, to exclude changes of the sepsis associated encephalopathy. To verify possible morphological differences between the yellow fever cases groups, between the control groups, and between the cases and the controls, we applied the statistical tests Fisher's exact test and chi-square, with p values < 0.05 considered statistically significant.

Results: All cases and controls presented, at least focally, neuropathological changes, which included edema, meningeal and parenchymal inflammatory infiltrate and hemorrhages, and perivascular inflammatory infiltrate. We did not find an unequivocal aspect of encephalitis. The only parameter that, after statistical analysis, can be attributed to yellow fever was the perivascular inflammatory infiltrate.

Conclusions: The neuropathological findings are sufficient to justify the multiple clinical neurologic disturbances detected in the YF cases. Since most of the parameters evaluated did not show statistically significant difference between cases and controls, an explanation for most of the neuropathological findings may be the vascular changes, consequent to shock induced endotheliopathy, associated with stimulation of the immune system inherent to systemic infectious processes. The statistical difference obtained in yellow fever cases regarding perivascular infiltrate can be can be explained by the immune activation inherent to the condition.