Uromodulin in sepsis and severe pneumonia - A two-sample Mendelian randomization study

Outcome for sepsis-associated acute kidney injury (AKI) patients in the intensive care unit (ICU) remains poor. Low serum uromodulin (sUMOD) protein levels have been proposed as a causal mediator of this effect. We investigated the effect of different levels of sUMOD on the risk of sepsis and severe pneumonia and outcomes in these conditions. A two-sample Mendelian Randomization (MR) study was performed. Single nucleotide polymorphisms (SNPs) associated with increased levels of sUMOD were identified and used as instrumental variables for association with outcomes. Data from different cohorts were combined based on disease severity and meta-analysed. Five SNPs associated with increased sUMOD levels were identified and tested in six data sets from two biobanks. There was no protective effect of increased levels of sUMOD on the risk of sepsis (two cohorts, OR 0.99 ( 95 % CI 0.95-1.03), p = 0.698 and OR 0.95 (0.91-1.00), p = 0.060 respectively), risk of sepsis requiring ICU admission (OR 1.04 (0.93-1.16), p = 0.467), ICU mortality in sepsis (OR 1.00 (0.74-1.37), p = 0.987), risk of pneumonia requiring ICU admission (OR 1.05 (0.98-1.14), p = 0.181) or ICU mortality in pneumonia (OR 1.17 (0.98-1.39), p = 0.079). Meta-analysis of hospital-admitted and ICU-admitted patients separately yielded similar results (OR 0.98 (0.95-1.01), p = 0.23 and OR 1.05 (0.99-1.12), p = 0.86 respectively). Among patients with sepsis and severe pneumonia there was no protective effect of different levels of sUMOD. Results were consistent regardless of geographic origins and not modified by disease severity.