Acute Kidney Injury: Diagnosis, Management and Controversies

Probably the area I feel least comfortable with on the AICU so looking forward to an update – especially with regard to RRT. Brexit jokes only allowed for first hour according to the chair….

Excellent overview from basic pathophysiology to RRT. 3 excellent reviews worth mentioning that were highlighted:

Acute kidney injury and sepsis:

Pathophysiology of acute ischaemic kidney injury:

Fluid management in AKI:

Acute Kidney Injury

Is all AKI the same? (Lui Forni)

NO! AKI is a syndrome –  IMPORTANTLY does not include the cause…


  1. Specific kidney diseases e.g. glomerulonephritis
  2. Non specific e.g. ischaemia
  3. Extra renal e.g. obstruction

Beware the hidden tiger in the AKI without obvious cause

Transjugular renal biopsy by interventional radiologists potential option in coagulopathic patient

Pathogenesis of AKI (John Prowle)

Again – reinforcing AKI is a heterogenous clinical syndrome – multifactorial

ATN first described in 1941 in Blitz survivors. Crush injuries with secondary mortality from renal failure:

Key points:

  1. Baseline risks e.g hypertension
  2. Acute pathology e.g. sepsis
  3. Nephrotoxicity e.g. drugs

Like playing the fruit machines in UK – if all three come up, highly likely to develop AKI

Treatment of the above:

  1. Recognise risks
  2. Treat and mitigate pathology
  3. Avoid unnecessary nephrotoxin exposure

Conventional belief that AKI is driven by global organ ischaemia – theory that has now been challenged

Commonest association with AKI is sepsis BUT unlikely due to purely global hypoperfusion

More important = balance between kidney ischaemia and inflammation injury

Endothelial cell activation, injury and reduced microvascular flow play major role – may not be dependent on global macrovascular flow –

Noradrenaline may increase glomerular pressure without compromising renal perfusion and restore GFR

BUT if tubular injury due to inflammation – cant be manipulated by systemic haemodynamics as reduction in renal blood flow in AKI – mediated by tubular dysfunction

Take home messages:

  1. Initially in septic shock – increasing BP may increase UO.
  2. In establised AKI – nil benefit. (except potentially in HT patients – renal hypertension challenge to assess improvement in UO)

Epidemiology of AKI – Ville Pettila

Incidence of AKI in ICU around 40-60%

RRT in AKI 300-600 per million population

Top 5 causes: Hypovolaemia/diuretics/hypotension/colloids/severe sepsis

Even stage 1 AKI associated with increased mortality

Interesting question about lack of routine followup for patients with AKI? e.g. compared to acute MI

Need to engage local nephrology services early in arranging followup review

Prevention of AKI – Miet Schetz

Reinforcing above: Patients at risk of AKI

  1. Baseline risks
  2. Acute conditions
  3. Nephrotoxicity

Interplay between risk factors v insults

Radiocontrast, aminoglycosides, ACE-I

Limited evidence in witholding ACE-I  – cardiac cath:

Seem to be functional but not structural changes to kidney – ?lack of evidence of harm

Fluid overload bad for the kidney. Two excellent papers on fluid management:

Choice of fluid also important:

General message – starches = bad

6S trial NEJM  –

Myburgh 2012 NEJM –

So then, saline or buffered crystalloid solution? – choride restrictive regime associated with reduced incidence of AKI and RRT – SPLIT trial – nil difference in renal outcomes with buffered crystalloid v saline (but ?low risk pts and low volumes crystalloids)

Take home messages:

  1. Starches nephrotoxic in sepsis and ICU
  2. Hyperchloraemia may induce renal vasoconstriction
  3. Hypotension duration/severity related to AKI
  4. Earlier use of vasopressor may be advantage
  5. No adverse effect of vasopressor on renal function in patients with vasoplegic hypotension
  6. Targets should be individualised


Renal replacement therapy

Case study? Need for RRT? (Eric Hoste)

Majority of people would not start RRT in AKI stage 1 in patient with abdominal sepsis and laparostomised abdomen

By AKI – stage 3, still a third of audience wouldnt start RRT

Such a disparity. Compared to cardiac/respiratory support – potentially clearer decisions on e.g. intubation etc

When to start RRT? (Michael Darmon)

Current practices

Survey of practice in nephrologists:

Median Ur 20/ Cr 322 for initiation of RRT

Should we start earlier in patients with haematological instability?  –  Small RCT – deleterious effect of haemofiltration during early phase sepsis

Risks/Benefits of early RRT pre-classical indications?

Risks of early RRT?

  1. Unrequired RRT in pt with spontaneous recovery
  2. Risk of catheter insertion and extracorporeal circuit
  3. Depletion of drugs, nutrients, electrolytes
  4. Resource consumption

Benefits of early RRT?

  1. Prevention of electrolye disturbance
  2. Homeostasis – better fluid balance

Should we start premptive RRT then? Before classical indications? – suggestion of beneficial effect on survival but heterogeneous studies ++

AKIKI study – RCT in NEJM 2016:

  • KDIGO stage 3
  • Within 6h or discouraged until usual criteria or >12h
  • No difference in mortality. Decrease in RRT requirement by 50%

ELAIN study may provide more data –


  • Definition of delayed RRT differs across studies
  • Most of studies are at high risk of bias
  • Factors leading to RRT initiation are not taken into account
  • A single study was randomised
  • Patients with severe AKI who did not require RRT are usually excluded from observational studies

Take home messages:

  1. Optimal timing remains unknown
  2. Most interventionalists would start RRT within 24 hours in KDIGO stage 3 patients (without waiting for conventional indications)
  3. Data resulting from RCTs remains conflicting and hard to interpret 
  4. Late strategy may help avoiding RRT in up to 50% patients
  5. What is the role of uraemia (beyond fluid overload) in systemic organ failure

What modality to choose? (Heleen Oudemans-van Straaten)

Key points from the talk:

  1. Timeframe (intermittent/continuous)
  2. Mechanism (HD/HF)
  3. Type of buffer (lactate v bicarbonate)
  4. Type of A/C (heparin v citrate v nothing)

Boils down to short term benefits, long term benefits and local expertise/facilites

Intermittent HD could provoke hypotension in unstable patient

Dialysis disequilibrium syndrome. Early from cirulation, late from brain —> cerebral oedema. Due to reversed osmotic shift

Cochrane review in 2007 – no difference in mortality/renal recovery and haemodynamic stability

BUT: Higher MAP and lower need for vasopressor dose escalation

KDIGO recommendation: Use IHD and CHD as complementary techniques in ICU patients

HD v HF v HDF – need to choose dependent on patient and timing

Take home messages:

  1. Initial RRT modality? Consider
    • Haem instability
    • Fluid overload
    • ICP
    • Renal recovery
    • IHD in patients with stable clinical picture and improving
  2. No robust clinical difference between HD and HF
  3. Cautious with lactate buffering in shock/liver dysfunction

Citrate v Heparin coming later!


Which membrane for RRT? (Thomas Rimmele)

Which polymer?

Molecule removal = function of size and plasma concentration.

But uraemic toxins and essential proteins – comparable mol weights

Plasma concentration – like looking for needle in a haystack? Might want to be more specific in which molecules to focus on