Renal Perfusion in Human Septic Shock

Student thesis: Doctoral ThesisDoctor of Philosophy


Sepsis is the dysregulated host response to infection and with resultant organ impairment carries a mortality rate of between 15 and 25%. It is a leading cause of acute hospital admissions. Septic shock, the severe end of the sepsis spectrum is identified by persistent lactataemia and vasoplegia.

Hypoperfusion distinguishes patients with this condition, its timely identification being critical to their management. Organ impairment is fundamental, with renal impairment being most frequently observed and independently associated with mortality. Despite its prevalence, the mechanisms underlying renal impairment in septic shock are unclear. Inflammation, metabolic alterations and cell cycle arrest are all implicated but alterations in renal perfusion are also contributory. Animal models are limited and the extent to which alterations in renal blood flow and perfusion occur in humans has not been determined.

Novel methods of studying renal perfusion are necessary as traditional cross-sectional imaging in early septic shock is difficult. The utility of bedside ultrasound in combination with dynamic contrast enhancement (DCE-US) is a potential tool to study these changes but has yet to be tested in detail. This thesis examines the extent to which renal perfusion is altered in both early and persistent septic shock and between patients who develop acute kidney injury and those who do not. It argues that these changes are fundamental to AKI development, anchoring them to biomarker profiles of renal and endothelial injury and inflammatory profiles.

Following a review of the background and a methodology of relevant techniques, Chapter 3 describes the development of the DCE-US technique and assesses its suitability for the study of critically ill patients, validating the developed method by the same user and between two users in a cohort of healthy controls. This study defines normal values of renal perfusion, previously unreported, and informative for the future assessment of patients. It describes my development of the outpatient radiology-based technique to one that is more suitable for bedside assessments in the intensive care unit (ICU), addresses reliability of contrast data analysis, by the same user (R=0.77-0.9) and between users (R=0.52-0.74) and develops a method of quantifying renal blood flow.

Chapter 5 compares the DCE-US variables in a patient study, addressing correlations between the variables and in combination with the healthy control data, compares and contrasts, aiming to gain greater insight into the strengths and weaknesses of the individual variables. The data presented here suggests a limitation of intensity-based measures. It introduces a grouping variable used throughout the thesis, that of renal perfusion status, dichotomized by those with above and below average values.

Chapter 6 describes the demographic data from the main patient study and compares baseline characteristics. The key differences are the higher vasopressor requirements and sickness severity in those who develop severe AKI. (day zero results: noradrenaline dose severe AKI 0.35(0.26-0.51)mcg/kg/min vs non-severe 0.21(0.14-0.3)mcg/kg/min, p<0.001; SOFA scores severe 11.3±3.28 vs non-severe 9.3±1.9, p<0.05).

Chapter 7 examines the primary outcome, that of renal microvascular alterations in sepsis. It compares renal perfusion according to the development of severe AKI and looks for longitudinal alterations in perfusion following a septic insult as patients either develop AKI or not. It demonstrates that patients with more pronounced impairment of renal cortical perfusion develop more severe AKI (day zero data mean transit time: severe AKI 10.2(6.5-23.2)sec vs non-severe 5.5(4.7-6.5)sec, p<0.05).

Chapter 8 examines haemodynamic data as renal perfusion and AKI manifests. It examines renal blood flow, measures of left heart function and cardiac output and right heart function with venous pressure and congestion assessment. This chapter demonstrates that renal perfusion alterations are intrinsic, generated by specific renal mechanisms and not secondary, captive to alterations in systemic haemodynamics.

Chapter 9 contrasts renal perfusion with systemic tissue perfusion and the wider examination of shock and its severity. It uses multiple assessments of global tissue perfusion, including sublingual incident dark-field microscopy and biochemical assessments. Further mechanistic data are presented from markers of endothelial and glycocalyx injury. This chapter provides insight into the distinct nature of renal hypoperfusion, its weak association with systemic alterations and its persistence, in contrast to the early resolution of global parameters.

Chapter 10 looks at the relationship between inflammation, renal perfusion and AKI development. It identifies an association with inflammation which occurs later during AKI and is associated with persistence (day 4 IL-8 values maintained perfusion group 32(15-62)ng/ml vs hypoperfusion perfusion group 69(45-154) ng/ml, p<0.05) and angiopoietin ratios by day 4 (maintained perfusion group 0.88(0.27-1.53) vs hypoperfusion group 2.11(1.86-5.82), p <0.05), these between group differences are not present on admission. Chapter 11 follows on from the previous chapter by examining renal perfusion in previously described AKI subphenotypes. These subphenotypes are differentiated by their inflammatory profiles and have differing outcomes following the use of vasopressors, suggesting a vascular aetiology.

Chapter 12 looks for correlations between renal biomarkers and perfusion, as novel markers are specific to the site of injury, associations may inform the areas of the nephron most affected by alterations in perfusion and longitudinal data demonstrates differing patterns over time, but the close association between tubular alterations and hypoperfusion. This chapter also examines the ability for perfusion assessment to predict severe AKI, demonstrating similar results when compared to novel biomarkers (AUROC perfusion index 0.84 vs AUROC NGAL*albuminuria 0.78).

Chapter 13 describes longer term data in relation to renal perfusion and whether hypoperfusion is more likely to result in persistent AKI and disease progression. It finds that hypoperfusion can predict the time spent on RRT and the development of acute kidney disease (admission mean transit time for those who spend <7 days on RRT 6.255(4.602-8.2) sec vs more 10.477(7.157-25.387) sec, p<0.05).

Chapter 14 compares alterations between septic shock and another severe infection without the septic phenotype, COVID-19 associated AKI. This provides a second cohort of critically ill patients to compare those with sepsis to and demonstrates the utility of DCE-US assessment.

This work describes the development and use of DCE-US as a novel tool, creating a window to assess renal status in sepsis associated AKI. This technique provides utility beyond that provided by novel biomarkers. Renal hypoperfusion appears to be fundamental to, and persistent in all patients with septic shock in combination with downregulation of tubular epithelial cells (TECs) but more so in those who develop the clinical manifestations of AKI. Patients with renal hypoperfusion have worse renal outcomes and a greater mortality. DCE-US accurately identifies these patients on admission and this thesis provides a foundation for future study.
Date of Award1 Oct 2023
Original languageEnglish
Awarding Institution
  • King's College London
SupervisorSam Hutchings (Supervisor) & Kate Bramham (Supervisor)

Cite this