Acute Kidney Injury After Cardiac Surgery


  • Cardiac surgery-associated AKI is one of the commonest complications of adult cardiac surgery
  • It is the second leading cause of AKI in ICU patients, after sepsis

  • Incidence is 5-40% depending on the definition used and the type of surgery:
    • Lowest after isolated CABG
    • Progressively higher after valve surgery, aortic surgery and heart transplantation
    • Up to 2% of patients will require renal replacement therapy
  • In reality, sub-clinical renal injury probably occurs in all patients undergoing cardiac surgery to some extent

  • Diagnosis is as per KDIGO criteria for standard AKI

Limitations of conventional markers

  • Creatinine:
    • Influenced by muscle mass and dietary protein intake, both of which vary perioperatively
    • Does not begin to rise until GFR is <50% of normal
    • This delay is lengthened further by renal functional reserve, i.e. compensatory hyperfiltration

  • Urine output:
    • Affected by fluid balance and diuretic use
    • Loses diagnostic value when clinicians intervene in response to oliguria

  • Cystatin C
    • Less dependent on muscle mass and diet
    • Rises rapidly with the onset of AKI
    • More expensive and less widely available

Novel biomarkers

  • Novel biomarkers aim to detect kidney injury earlier than creatinine, allowing earlier intervention
Biomarker Key points Limitations
TIMP-2 × IGFBP7
('NephroCheck')
Urinary cell-cycle arrest proteins from proximal tubular cells
First biomarkers available for clinical use
Predicts AKI as early as 4h post-op
Expensive
Not specific to cause of injury
Inversely correlated with urine output, so dilute urine may underestimate injury
NGAL Released by damaged tubular cells
Rises quickly, peaking ~6h post-op
Not kidney specific (also lung, liver)
Raised in inflammation, infection and CKD
KIM-1 Proximal tubular protein upregulated after ischaemia-reperfusion injury Slower rise and lower accuracy
Expensive
Mainly a research tool
DKK3 Pre-operative urinary level predicts post-op AKI
May identify those at risk of progression to CKD
Research only
Limited validation

Other techniques under investigation

  • Renal functional reserve: the rise in GFR in response to an IV amino acid load; lower in patients who go on to develop AKI
  • Bladder urine oxygen tension: a surrogate for renal medullary oxygenation; lower intra-operatively in those who develop AKI, but unreliable in anuria
  • Renal resistive index: measured with TOE; a raised index is associated with higher AKI incidence

  • In general, anything that reduces renal oxygen delivery or diminishes functional renal reserve increases risk

Patient risk factors

  • Older age
  • Pre-existing CKD (prevalence ~20% of those presenting for cardiac surgery)
  • Comorbidities which reduce renal DO2:
    • Anaemia
    • Pulmonary disease e.g. COPD
    • Heart failure/low ejection fraction
  • Diabetes
  • Hypertension
  • Raised BMI
  • Non-White ethnicity
  • Female sex has traditionally been considered a risk factor, but this may simply reflect women presenting older and with more comorbidity

Surgical risk factors

  • Prior cardiac surgery
  • Emergency surgery
  • Type of surgery: valve, aortic and transplant surgery carry higher risk than CABG
  • Long duration of surgery, CPB and aortic cross-clamp
  • Low flow phases during CPB
  • Hypoperfusion or hypovolaemia during or after surgery
  • Venous congestion
  • Vasoactive drug use
  • Blood transfusion
  • Need for IABP

Prediction models

  • Several exist, including the Cleveland Clinic model, Mehta score, Simplified Renal Index and Leicester score
  • Four variables feature consistently:
    • Pre-operative renal function
    • Diabetes
    • Previous cardiac surgery
    • Type of cardiac surgery
  • The Leicester score is the first to predict any stage of AKI, including stage 1
  • Rarely used clinically, likely because there is no clear intervention to follow a high-risk result


Associations of AKI in cardiac surgery
↑ short-term morbidity e.g. infection, MI, stroke, heart failure
↑ short-term mortality; if requiring dialysis mortality is 50-60%
Five-fold ↑ risk of developing CKD; AKI may play a direct causal role
↑ length of ICU & hospital stay
↑ likelihood of discharge to an extended care facility
↑ long-term mortality, extending up to 10 years after surgery
↑ economic cost


  • Complex and multifactorial with self-perpetuating cycles, but converges on a few key mechanisms:
    • Hypoperfusion
    • Hypoxia
    • Neurohumoral activation
    • Inflammation
    • Oxidative stress
  • These mechanisms are also the targets of preventive strategies

Hypoperfusion

  • Reduced cardiac output and hypotension reduce renal blood flow (RBF), particularly to the medulla
  • Autoregulation (myogenic response and tubuloglomerular feedback) is impaired during cardiac surgery due to:
    • MAP falling below the autoregulatory threshold (~80mmHg)
    • Hypothermia
    • Haemodilution
    • Non-pulsatile CPB flow

  • The kidney's two capillary beds in series require a higher driving pressure than other organs
  • Blood is preferentially redistributed away from the kidneys to the heart and brain during CPB
  • CPB-induced haemolysis lowers nitric oxide (NO) levels yet the medulla relies heavily on NO for vasodilation
  • Venous congestion also reduces GFR, in proportion to the rise in renal venous pressure

Hypoxia

  • The kidneys receive ~20% of cardiac output, but oxygen is unevenly distributed:
    • Cortical PO2 ~50mmHg
    • Medullary PO2 ~10-30mmHg

  • Active Na+ reabsorption accounts for ~80% of renal oxygen consumption
  • Arteriovenous oxygen shunting in the cortex limits DO2 to the outer medulla, predisposing it to hypoxia
  • Reduced RBF plus haemodilution from CPB priming reduces renal DO2 by ~20%
  • Stress hyperglycaemia increases oxygen consumption via increased Na+-glucose co-transport in the proximal tubule
  • Renal hypoxia may persist for weeks after CPB, potentially contributing to fibrosis

Neurohumoral response

  • Hypotension activates the sympathetic nervous system via baroreceptors
  • Renal noradrenaline release causes α-mediated afferent and efferent arteriolar vasoconstriction, reducing RBF
  • Noradrenaline also stimulates renin release
  • Net effect: GFR is maintained but RBF falls, lowering cortical oxygen tension

Inflammation

  • Triggered by damage-associated molecular patterns (DAMPs) released from injured tissue and blood contact with the CPB circuit
  • Complement is activated by hypoxia, DAMPs, the CPB circuit and protamine
  • This leads to cytokine release, leucocyte activation, vasodilation and increased capillary permeability
  • Reperfusion brings leucocyte infiltration into injured tissue: ischaemia-reperfusion injury

Oxidative stress

  • Succinate accumulates during hypoxia and is oxidised on reperfusion to generate reactive oxygen species (ROS)
  • ROS cause mitochondrial damage and further DAMP release
  • CPB-induced haemolysis releases free haemoglobin and iron, which:
    • Increase oxidative stress
    • Scavenge NO, causing renal vasoconstriction
    • Form haem casts, causing tubular obstruction

  • Many interventions have been studied but most have proven ineffective
  • Until new evidence emerges, the KDIGO bundle of care remains the recommended approach

General measures

  • Identify risk factors and modify where possible
  • Screen for and treat pre-operative anaemia: iron ± erythropoietin in selected patients
  • Minimise blood transfusion
  • Avoid rapid rewarming (no faster than 0.5°C/min) and hyperthermia
  • Arterial outflow temperature >37°C on CPB is associated with AKI
  • Off-pump surgery reduces AKI incidence in meta-analyses, though its use is limited by surgical and institutional factors

Promising strategies

  • IV amino acids:
    • Reduce afferent arteriolar resistance and increase NO activity, raising RBF and GFR
    • The PROTECTION trial showed a reduction in AKI (26.9% vs 31.7%) with a continuous infusion from the start of surgery for up to 72h
    • Unclear whether this reflects true reduction in structural kidney injury

  • Vasopressin:
    • Constricts the efferent more than the afferent arteriole, potentially preserving GFR
    • Associated with lower AKI incidence than noradrenaline in vasoplegic shock after cardiac surgery

  • Goal-directed perfusion: targeting DO2 thresholds on CPB reduced AKI in meta-analysis, but targets are not yet standardised
  • SGLT2 inhibitors: may reduce renal oxygen consumption and promising pilot data with large trials ongoing
  • Extracorporeal cytokine filtration during CPB: one trial showed reduced AKI

Not shown to be effective

  • Higher MAP targets: no reduction in AKI
  • Remote ischaemic preconditioning: no benefit in two large RCTs
  • Furosemide: intended to reduce medullary oxygen consumption but abandoned after negative trials
  • 'Renal dose' dopamine: increases RBF and urine output but offers no renal protection and may worsen tubular injury
  • Mannitol: not robustly shown to prevent AKI
  • Corticosteroids: no benefit in two large RCTs
  • Dexmedetomidine: inconsistent data, and associated with more hypotension
  • Sodium bicarbonate: no reduction in AKI, RRT, ICU stay or mortality
  • Vitamin C: no benefit at conventional doses
  • NO supplementation via the CPB circuit: small studies suggest benefit, but concern regarding subclinical injury

KDIGO bundle of care

  • Targets multiple pathological mechanisms at once
  • In the PrevAKI-multicentre trial, applying the bundle to high-risk patients reduced moderate-severe AKI (14% vs 24%)
Recommendation Detail
Avoid nephrotoxins Withhold ACE inhibitors and ARBs for 48h post-op
Avoid NSAIDs, aminoglycosides and contrast
Optimise volume status and perfusion pressure Advanced haemodynamic monitoring
MAP ≥65mmHg
Close monitoring Serum creatinine and urine output to diagnose AKI early
Avoid hyperglycaemia Target glucose 6.1 - 8.3 mmol/L, with insulin if required

Established AKI

  • Management is supportive and similar to AKI of any aetiology
  • Continue the bundle measures above
  • Treat complications e.g. hyperkalaemia, acidosis, fluid overload
  • Consider RRT in those with refractory issues