Rationale: The global burden of sepsis is concentrated in sub-Saharan Africa, where inciting pathogens are diverse and HIV co-infection is a major driver of poor outcomes. Biological heterogeneity inherent to sepsis in this setting is poorly defined.

Objectives: To identify dominant pathobiological signatures of sepsis in sub-Saharan Africa and their relationship to clinical phenotypes, patient outcomes, and biological classifications of sepsis identified in high-income countries (HICs).

Methods: We analyzed two prospective cohorts of adults hospitalized with sepsis (severe infection with qSOFA score≥1) at disparate settings in Uganda (discovery cohort [Entebbe, urban], N=242; validation cohort [Tororo, rural], N=253). To identify pathobiological signatures in the discovery cohort, we applied unsupervised clustering to 173 soluble proteins reflecting key domains of the host response to severe infection. A random forest-derived classifier was used to predict signature assignment in the validation cohort.

Measurements and Main Results: Two signatures (Uganda Sepsis Signature [USS]-1 and USS-2) were identified in the discovery cohort, distinguished by expression of proteins involved in myeloid cell and inflammasome activation, T cell co-stimulation and exhaustion, and endothelial barrier dysfunction. A five-protein classifier (AUROC 0.97) reproduced two signatures in the validation cohort with similar biological profiles. In both cohorts, USS-2 mapped to a more severe clinical phenotype associated with HIV and related immunosuppression, severe tuberculosis, and increased risk of 30-day mortality. Substantial biological overlap was observed between USS-2 and hyperinflammatory and reactive sepsis phenotypes identified in HICs.  

Conclusions: We identified prognostically-enriched pathobiological signatures among sepsis patients with diverse infections and high HIV prevalence in Uganda. Globally inclusive investigations are needed to define generalizable and context-specific mechanisms of sepsis pathobiology, with the goal of improving access to precision medicine treatment strategies.

In cryopreserved serum samples, Olink proteomics (Olink Proteomics AB, Uppsala, Sweden) was performed at the Human Immune Monitoring Center of the Icahn School of Medicine at Mount Sinai (New York, NY, USA) using Target Immunooncology and Cardiometabolic panels. These panels, each of which include 92 proteins, were selected to broadly capture pathobiological domains implicated in the host response to severe infection (i.e., innate and adaptive immune activation and exhaustion, endothelial and cellular metabolic dysfunction, dysregulated coagulation). Comprehensive descriptions of each panel including validation data are available at https://olink.com/products-services/target/immune-response-panel/ and https://olink.com/products-services/target/cardiometabolic-panel/. Samples from RESERVE-U-1-EBB and RESERVE-U-2-TOR were analyzed separately. For each cohort, samples were analyzed in a single batch, randomized across plates, and analyzed by technicians blinded to clinical data. For each panel, proteins were quantified via the Olink Proximity Extension Assay. The Proximity Extension Assay uses dual oligonucleotide-labelled antibodies to bind target proteins, creating a unique double stranded DNA barcode which is quantitatively proportional to the initial concentration of the target protein. This barcode then serves as a template for a DNA polymerase-dependent extension step, which is followed by PCR amplification (T100 Thermal Cycle, Bio-Rad Laboratories, Inc., Hercules, CA, USA). Resulting DNA amplicons are quantified by microfluidic qPCR (Fluidigm Biomark HD and JUNO Systems, Standard BioTools Inc., San Francisco, CA, USA) and normalized across batch plates. Intensity normalization was used for RESERVE-U-1-EBB assays and reference sample normalization was used for RESERVE-U-2-TOR assays. Relative protein abundance was calculated from cycle threshold values and expressed in log₂-normalized protein expression (NPX) units by Olink’s NPX Manager software. Missing NPX values, which were rare, were multiply imputed for each panel using chained equations, predictive mean matching, and all other proteins as predictors (mice R package). Five imputed datasets were reviewed for convergence and plausibility after which one was randomly selected for use.