Abstract
Eculizumab significantly increased survival in patients with transplant associated thrombotic microangiopathy (TA-TMA), however one third of TA-TMA patients still die despite prompt treatment with eculizumab. The mechanism of eculizumab resistance in TA-TMA is unknown and no prior study investigated the contribution of circulating leukocytes to TA-TMA biology. We hypothesize that cellular and soluble complement pathway modifiers contribute to eculizumab resistance in TA-TMA and knowledge of these mechanisms may guide novel biomarkers and therapies.
To test this hypothesis, we first performed a pharmacokinetic (PK)/pharmacodynamic (PD) analysis of eculizumab drug levels and sC5b-9 levels in 40 eculizumab responders and 20 eculizumab non-responders. Peripheral blood mononuclear cells (PBMCs) from a subset of responders (n=4) and non-responders (n=4) then underwent single cell RNAseq at two timepoints: 1) TA-TMA diagnosis (prior to eculizumab initiation) and 2) at the time of eculizumab “escape” in non-responders (or matched timepoint in responders). The “escape” timepoint was determined using individualized PK/PD data. Differentially expressed genes (DEGs) were entered into Qiagen Ingenuity Pathway Analysis software for pathway and upstream regulator analysis.
PK/PD studies showed non-responders had higher eculizumab clearance (p<0.001) and were less likely to achieve target trough levels (p=0.03). Non-responders also had higher sC5b-9 levels at diagnosis (p=0.03) and had higher maximum sC5b-9 levels during the treatment course (p=0.001). Additionally, non-responders were more likely to develop gastrointestinal bleeding (p<0.001), had higher maximum urine protein to creatinine ratios (p=0.008) and worse overall survival (p=0.001).
Single cell RNAseq identified 12 cell types in patient PBMCs for comparison. First, we compared non-responder transcriptomes to responder transcriptomes at the time of TA-TMA diagnosis. Myeloid cell abundance was similar in classical monocytes (42% vs 48%), non-classical monocytes (4.3% vs 5.3%) and dendritic cells (13.4% vs 11.9%) between non-responders and responders, however non-responder myeloid cells showed a consistent pattern of interferon pathway enrichment. The top 3 enriched pathways in classical monocytes were interferon alpha/beta signaling (p=1.3e-12), interferon signaling (p=3.4e-8) and interferon gamma signaling (p=5.6e-8). Complement system activation was also observed (p=7.2e-5). Non-classical monocytes were similarly enriched for interferon alpha/beta signaling (p=7.3e-9) and interferon signaling (p=3.9e-5). Lastly, dendritic cells were enriched for interferon gamma signaling (p=2.5e-5) and interferon alpha/beta signaling (p=4.2e-6). This suggests interferon production from donor monocytes and dendritic cells may contribute to a unique TA-TMA phenotype in non-responders.
Non-classical monocytes were notably enriched for multiple senescence pathways including DNA damage/telomere stress induced senescence (p=2.8e-9), senescence-associated secretory phenotype (p=3e-4) and oxidative stress induced senescence (p=4.3e-3). Cell senescence is induced by interferons and in this scenario may impede the crucial role of non-classical monocytes in vascular maintenance and repair. This is further supported in our comparison of non-responder transcriptomes to responder transcriptomes at the time of eculizumab escape. Non-classical monocytes in non-responders at the time of eculizumab escape showed strongly inhibited eukaryotic translation elongation (p=2.9e-33), EIF2 signaling (p=1.5e-30), eukaryotic translation initiation (p=1.6e-27) and RNA processing pathways (p=2.2e-25). These findings suggest highly impaired cell translation and overall function has occurred in non-classical monocytes at the time of eculizumab escape.
In conclusion, we performed the first PK/PD guided single cell RNAseq analysis of PBMCs in eculizumab responders and non-responders. This study identified the plausible contribution of interferons to eculizumab resistance in non-responders. Additionally, we identified a potentially novel mechanism of impaired endothelial cell repair in TA-TMA through interferon-induced senescence in non-classical monocytes. Paired plasma proteomic samples from the patients in this study are currently being analyzed for the purpose of validating these observations and identifying candidate targets for biomarker and treatment studies.
