Associations of inflammatory markers with post-acute clinical findings among survivors of Ebola virus disease with and without viral RNA shedding in the semen in Liberia: a nested case-control study.

backgroundA high proportion of survivors of Ebola virus disease (EVD) have post-acute sequelae of EVD (PASE), but the relationship between inflammation and PASE pathogenesis is poorly understood. This study tests the hypothesis that inflammation is associated with PASE among survivors with and without viral RNA shedding in the semen.

methodsThis was a case-control study nested in a longitudinal cohort that recruited confirmed survivors of EVD and their uninfected contacts from the 2013-16 EVD epidemic in Liberia, starting on June 1, 2015. We included participants aged at least 18 years with clinical data and plasma available at cohort baseline for analysis. A semen donation substudy tested male survivors for Ebola virus RNA shedding in the semen. A sex-stratified and survivor-stratified random sample of cases (survivors) and controls (contacts) was obtained to select stored baseline plasma samples for cytokine testing of markers of inflammation, immune regulation, and antiviral responses. Serostatus of cases and controls was confirmed by Filovirus Animal Nonclinical Group assay. We identified inflammatory markers (adjusted p≤0·05) elevated in cases compared with controls and then used these biomarkers in analyses comparing survivors with and without pre-specified PASE-associated clinical findings (self-reported symptoms and abnormal examination findings). Survivors with viral RNA shedding in the semen formed subgroup analyses.

findingsOur analysis cohort consisted of 1044 participants (594 survivors of EVD and 450 uninfected contacts); 515 (49·3%) were female and 529 (50·7%) were male. The subcohort of 243 male survivors with data on viral shedding included 81 (33%) participants with viral shedding in semen. Median time from acute EVD to baseline was 317 days (IQR 271-366). Survivors of EVD showed a pattern of elevated inflammatory markers indicative of macrophage (MCP-1, IL-1β, and M-CSF) and angiogenic factor activation (VEGF-A) compared with controls (adjusted p<0·05). In survivors with viral shedding in the semen compared with controls, VEGF-A was the only inflammatory marker that was significantly higher (adjusted p<0·001). After restricting the analysis to survivors, each inflammatory marker had a specific pattern of clinical findings. Higher levels of IL-1β were associated with higher odds of urinary frequency (p=0·002), musculoskeletal abnormalities (p=0·003), and abdominal abnormalities (p=0·03). By contrast, higher levels of MCP-1 were associated with lower odds of the same clinical findings. M-CSF was the only inflammatory marker associated with lower odds of joint pain (p=0·04). Higher levels of VEGF-A were associated with higher odds of abnormal chest findings in the overall survivor group (p=0·02) and in the subgroup with viral shedding in the semen (p=0·02).

interpretationWe found evidence of distinct biological pathways for PASE. Although viral RNA shedding in the semen could be associated with angiogenic activation, it did not explain many of the PASE symptoms and exam findings associated with the elevated macrophage markers, suggesting the pathobiology of some clinical manifestations might be autoimmunity, immune dysregulation, or another biological mechanism. These findings could inform shared biological pathways with other infection-associated chronic conditions, including post-acute sequelae of SARS-CoV-2 infection.

fundingNational Cancer Institute and National Institute of Allergy and Infectious Diseases at the US National Institutes of Health.