Identification of biomarkers for COVID-19 associated secondary hemophagocytic lymphohistiocytosis

Abstract

Objectives: We aimed to define and validate novel biomarkers that could identify individuals with COVID-19 associated secondary hemophagocytic lymphohistiocytosis (sHLH) and to test whether fatalities due to COVID-19 in the presence of sHLH were associated with specific defects in the immune system.

Design: In two cohorts of adult patients presenting with COVID-19 in 2020 and 2021, clinical lab values and serum proteomics were assessed. Subjects identified as having sHLH were compared to those with COVID-19 without sHLH. Eight deceased patients defined as COVID-sHLH underwent genomic sequencing in order to identify variants in immune-related genes.

Setting: Two tertiary care hospitals in Seattle, Washington (Virginia Mason Medical Center and Harborview Medical Center).

Patients: 186 patients with COVID-19.

Interventions: None.

Measurements and main results: Nine percent of enrolled COVID-19 subjects met our defined criteria for sHLH. Using broad serum proteomic approaches (O-link and SomaScan), we identified three biomarkers for COVID-19 associated sHLH (soluble PD-L1, TNF-R1, and IL-18BP), supporting a role for proteins previously associated with other forms of sHLH (IL-18BP and sTNF-R1). We also identified novel biomarkers and pathways of COVID-sHLH, including sPD-L1 and the syntaxin pathway. We detected variants in several genes involved in immune responses in individuals with COVID-sHLH, including in DOCK8 and in TMPRSS15, suggesting that genetic alterations in immune-related genes may contribute to hyperinflammation and fatal outcomes in COVID-19.

Conclusions: Biomarkers of COVID-19 associated sHLH, such as soluble PD-L1, and pathways, such as the syntaxin pathway, and variants in immune genes in these individuals, suggest critical roles for the immune response in driving sHLH in the context of COVID-19.

Figure 1.. sPD-L1, sTNF-R1, and IL-18BP are elevated two independent cohorts of COVID-sHLH subjects.

(A) Plasma samples (n=85) from cohort 1 were analyzed on the Olink platform and displayed as intensity normalized values. A.U. indicates arbitrary unit. Open circles: no sHLH (n=78, except for PD-L1 and TNF n=77), red squares: with sHLH (n=7). (B) Plasma samples from cohort 2 were analyzed using the SomaScan platform (n=101) and displayed as normalized RFU. Open circles: no sHLH (n=91); red squares: with sHLH (n=10). (C) Plasma samples from cohort 2 were analyzed for sPD-L1 (n=181) using an MSD assay. Open circles: no sHLH, red squares (n=164): with sHLH (n=17). Bars indicate median and interquartile range. * p < 0.05, ** p< 0.01 by two-tailed Mann-Whitney test.

Figure 2.. Syntaxin and IL-1 pathway proteins and CHCHD10 are differentially expressed in COVID-sHLH subjects.

(A) Plasma samples from cohort 2 were analyzed using SomaScan platform (n=101; no sHLH n=91; sHLH n=10). Volcano plot displays differentially expressed proteins between subjects with sHLH vs those without sHLH adjusted for age, sex, and BMI. FDR < 0.05. (B) Plasma levels of CHCHD10 from cohort 2 displayed as normalized RFU from SomaScan data presented in Panel A. Bars indicate median and interquartile range. (C) Normalized read counts of transcripts from RNA sequencing of CD14+ monocytes from subjects without sHLH (n=22) or subjects with sHLH (n=3) from cohort 1. Open circles: no sHLH, red squares: with sHLH. Bars indicate median and interquartile range. *** p < 0.001 by two-tailed Mann-Whitney test.