Abstract
Normal kidney function is essential to maintain whole body homeostasis. An acute decline in kidney function, ‘acute kidney injury’ (AKI), is in and by itself a major cause of perioperative morbidity and mortality. Maintaining preexisting kidney function therefore is a key task of the anaesthesiologist in the perioperative period [1].
Studies on perioperative AKI (including worsening of chronic renal failure) have mainly focused on the postoperative/ICU setting because (1) renal dysfunction does not alter intraoperative haemodynamics or oxygenation (provided a neutral fluid balance is maintained); (2) we lack readily available biomarkers to monitor intraoperative renal function (intraoperative oliguria is a poor marker of AKI, and creatinine value takes hours to rise); (3) medical treatment of AKI is mainly performed in the ICU (managing fluid overload, hyperkalaemia, drug dosing adjustments or renal replacement therapies); (4) patients at risk for AKI are likely to be admitted to the ICU postoperatively.
Introduction
Normal kidney function is essential to maintain whole body homeostasis. An acute decline in kidney function, ‘acute kidney injury’ (AKI), is in and by itself a major cause of perioperative morbidity and mortality. Maintaining preexisting kidney function therefore is a key task of the anaesthesiologist in the perioperative period [1].
Studies on perioperative AKI (including worsening of chronic renal failure) have mainly focused on the postoperative/ICU setting because (1) renal dysfunction does not alter intraoperative haemodynamics or oxygenation (provided a neutral fluid balance is maintained); (2) we lack readily available biomarkers to monitor intraoperative renal function (intraoperative oliguria is a poor marker of AKI, and creatinine value takes hours to rise); (3) medical treatment of AKI is mainly performed in the ICU (managing fluid overload, hyperkalaemia, drug dosing adjustments or renal replacement therapies); (4) patients at risk for AKI are likely to be admitted to the ICU postoperatively.
Despite being ‘silent’ in the operating room, measures to prevent postoperative AKI must be implemented pre- and intraoperatively and include identifying risk factors related to the type of surgery and the patient (preoperative condition), carefully managing fluid balances (both fluid overload and hypovolaemia should be avoided) and carefully considering the indication for nephrotoxic drugs.
Epidemiology and Definition of Acute Kidney Injury
Definition and Incidence
In 2004, the Acute Dialysis Quality Initiative reached a consensus definition of AKI based on the RIFLE criteria (see below) [2]. RIFLE is the acronym of Risk, Injury, Failure, Loss and End-Stage Kidney. It provides a structured classification of AKI severity and recovery. Prior to this consensus, the use of inconsistent definitions caused the incidence of AKI to widely range from 1% to 31%. The KDIGO (Kidney Disease: Improving Global Outcomes) guidelines define AKI as:
1. An increase in serum creatinine ≥ 0.3 mg/dL within the last 48 h
2. A 1.5 increase in baseline serum creatinine, known or presumed to have occurred within the last seven days
3. Urine output < 0.5 mL/kg/h for six hours.
When this KDIGO criteria were applied in a meta-analysis of 312 studies [3], the pooled incidence of AKI during an episode of hospital care was 21.6% in adults and 33.7% in children. It was highest in critical care patients (31.7%) and in patients who underwent cardiac surgery (24.3%). AKI-associated mortality was 23.9% in adults and 13.8% in children.
AKI is further stratified in three stages (see Table 16.1). Mortality increases with each stage.
| Stage | Serum Creatinine Increase | Urine Output |
|---|---|---|
| 1 |
| < 0.5 mL/kg/h for 6-12 h |
| 2 | 2–2.9 times baseline | < 0.5 mL/kg/h for ≥12 h |
| 3 |
| < 0.3 mL/kg/h for ≥ 24 h or anuria for ≥ 12 hours |
Special Considerations in the Perioperative Period
While the diagnosis of postoperative AKI obviously considers urine output as well, it has to be recognized that oliguria in the postoperative period is often secondary to the normal physiological retention of salt and water in response to tissue damage, pain, mild degrees of hypovolaemia, hypotension and positive pressure ventilation [4]. Few prospective studies have examined the contribution of oliguria to postoperative prognosis, in particular the ability of oliguria to predict subsequent creatinine changes. A study of critically ill surgical and medical patients reported that oliguria was not a useful predictor of subsequent increases in creatinine, and that there was no consistent relationship between the duration of oliguria and RIFLE criteria [5]. Another study confirmed that intraoperative oliguria is common but not usually followed by a rise in creatinine, with the authors suggesting that the relationship between oliguria and renal failure should be further investigated [6].
Risk Factors for Perioperative Acute Kidney Injury
In a review of 28 studies involving 10,865 patients that underwent either vascular, cardiac, general or biliary surgery, preexisting chronic renal failure was the most important and consistent risk factor for postoperative AKI [7]. As mentioned above, the incidence of AKI during an episode of hospital care was highest in critical care patients (31.7%) and in patients that underwent cardiac surgery (24.3%). In a single centre prospective study involving 15,102 patients undergoing non-cardiac surgery and without preexisting renal dysfunction (creatinine clearance > 80 mL/min), the incidence of postoperative renal failure was 0.8%, and was associated with the following seven independent preoperative risk factors [6]: age > 59, emergency surgery (as defined by ASA physical status), chronic liver disease, body mass index > 32, peripheral vascular disease, COPD requiring bronchodilator therapy and finally high risk surgery: intrathoracic, intraperitoneal, supra-inguinal vascular and other surgeries with a potential for large fluid shifts such as multilevel spine fusions, intracranial aneurysm clippings, transhiatal oesophagectomy and pelvic exenteration.
Pathophysiology
Renal Autoregulation
While a detailed description of renal physiology is outside the scope of this chapter, some basic concepts should be reviewed. Traditional teaching holds that autoregulation maintains renal blood flow in response to changes in mean arterial pressure within the 50 to 150 mmHg range in normotensive patients by vasodilating and vasoconstricting the afferent arteriole in response to a decrease and an increase of blood pressure, respectively. In the hypertensive patient, the curve is shifted to the right. The renal perfusion pressure is calculated as the difference between mean arterial pressure and central venous pressure. The perfusion pressure needs to remain above the lower autoregulation limit in order to maintain renal blood flow and hence glomerular filtration rate – once the perfusion pressure drops below the lower autoregulation limit, renal blood flow decreases proportionally with perfusion pressure.
While the above is true in the experimental environment in the isolated kidney, the intact kidney is extensively innervated by the sympathetic nervous system and influenced by systemic and locally released hormones and vasoactive substances. In fact, the response to circulatory failure (low blood pressure) consists of activation of the sympathetic nervous system and the renin axis, which results in afferent arteriolar vasoconstriction that maintains renal perfusion pressure [8]. Lack or exhaustion of this compensatory response (typical in distributive shock due to e.g. sepsis, high risk surgery, anaphylaxis) leads to systemic hypotension with kidney ischaemia despite the high cardiac output. In this context, noradrenaline has been demonstrated not to cause deterioration in kidney function [9], and for now is considered the best intraoperative approach to restore systolic blood pressure to minimize the incidence of AKI if combined with the judicious use of fluid therapy avoiding fluid overload [10].
Classification of Acute Kidney Injury and Limitations in Perioperative Care
Of all AKI cases, one-third are related to previous surgical procedures [4]. The diagnosis of AKI and its aetiology must be performed based on clinical criteria supported by ultrasound studies, biochemistry and urinalysis (urea/creatinine ratio, sodium, fractional sodium and urea excretion). The aetiology can be pre-, intra-, and post-renal (see Table 16.2), a distinction that helps guide treatment and prognosis.
Table 16.2. Potential causes of perioperative AKI. NSAIDs (Non-steroidal anti-inflammatory drugs), ACEI (Angiotensin converting enzyme inhibitors), ARB (Angiotensin receptor blockers).
| Pre-renal |
|---|
|
| Renal |
|---|
|
| Post-renal |
|---|
|
While they rarely are the origin of renal dysfunction (1–2%), post-renal causes can be easily diagnosed and thus have to be excluded first. They are most often marked by sudden oliguria or anuria. Intraoperatively, sudden oliguria or anuria has to prompt a search for a kinked Foley catheter. An ultrasound exam may reveal other causes in the perioperative period, e.g. renal vascular occlusion.
After having excluded post-renal causes, pre-renal and intra-renal causes have to be considered. This distinction is important because most cases of AKI during the perioperative period (70% in the recent EPI-AKI study [11]) are considered to be caused by renal ischaemia secondary to haemodynamic derangements (haemorrhage, perioperative losses, sepsis, low cardiac output). The effect of renal ischaemia goes through two phases. In the first phase, there are no structural changes, and renal function will recover rapidly (pre-renal or functional) after restoring perfusion. In the second phase, persistent hypoperfusion will cause structural renal damage, the deterioration of renal function will be more protracted, and prognosis will be worse.
Recent data suggest that the commonly used biochemical and urinalyses are not very useful in differentiating pre- and intra-renal causes in the context of sepsis and the critically ill surgical patient [12, 13]. In addition, renal hypoperfusion may involve more than just haemodynamic factors [14], and renal hypoperfusion per se may not be the leading cause of AKI in the critically ill and high risk surgery patient (defined as a patient with high risk of worsening or developing organ dysfunction postoperatively). There is evidence that fluid overload should be avoided because it might cause and propagate AKI [12, 15, 16], but also that aggressive fluid restriction should be avoided [16b].
Biomarkers of Acute Kidney Injury
In the last years, much research has been devoted to the validation of new biomarkers for AKI to try to better characterize both the clinical syndrome and its pathophysiological course [17]. They vary in their anatomical and cellular origin, physiological function, time of release after the onset of renal injury and kinetics after their release.
The main objective of a biomarker in this context would be to allow an early diagnosis of AKI, predict evolution and prognosis, and define basic pathophysiological mechanisms. Biomarkers for AKI can be stratified according to the part of the nephron that has been affected and/or the specific markers released by these sites if injured.
The availability of these new markers has led to a new classification of AKI based on the alteration of glomerular filtration (marked by creatinine levels) and the presence of structural damage (marked by elevated biomarkers). The complex clinical syndrome of AKI can thus be stratified in three new classes with prognostic significance:
– Subclinical AKI: patients with elevated biomarkers but no elevation of serum creatinine.
– Functional AKI: patients with elevated serum creatinine but no elevation of biomarkers.
– Structural AKI: patients with both elevated serum creatinine and biomarkers.
A New Acute Kidney Injury Paradigm for Sepsis and its Applicability to the High Risk Surgical Patient
As mentioned above, sepsis is considered to be a cause of (or a major factor contributing to) AKI in the critically ill patient [11, 18], with absolute or relative hypovolaemia and the concomitant use of vasoconstrictors resulting in ischaemic kidney injury. Even though physiologically plausible, recent experimental and clinical data challenge the vision that both septic and non-septic AKI are caused by ischaemia:
AKI is not a universal outcome after cardiac arrest. Cardiac arrest is the best model of clinical ischaemia with periods of low or absent renal flow, especially if there is no hypoperfusion after return of spontaneous circulation [12].
During sepsis, the proportion of cardiac output received by the kidneys does not change, implying that renal blood flow is high in patients with a hyperdynamic circulation. A hyperdynamic circulation is the most frequent form of haemodynamic derangement during sepsis [19].
AKI during sepsis can develop without any clinical evidence of hypovolaemia and haemodynamic instability [20].
Pathological findings in kidneys of patients dying from septic shock demonstrate heterogeneous tubular damage with apical vacuolization without extensive necrosis rather than acute tubular necrosis, the pathological correlate of ischaemic renal damage [21].
Based upon clinical data and robust experimental models, the following theory has been developed to explain the aetiology of sepsis induced AKI [22]. During early sepsis, the predominant mechanism of kidney dysfunction is a vasodilation state causing low glomerular filtration pressures and low glomerular filtration rates, despite the hyperdynamic circulation with a high cardiac output and high renal blood flow. In this early phase, treatment with vasoconstrictors can prevent AKI, but sometimes restoring haemodynamics will not be enough to restore kidney function because microcirculatory glomerular and peritubular dysfunction will inflict inflammatory damage. Filtration of the resulting inflammatory mediators will induce endothelial dysfunction that contributes to a sluggish peritubular blood flow that promotes adhesion of activated leucocytes to the endothelium that will infiltrate the kidney interstitium. The resulting inflammation will initially only affect proximal tubular cells because these are the cells first to be exposed to the inflammatory milieu. Next, these proximal tubular cells will secrete paracrine factors that trigger more distal tubular cells to enter into a sort of hibernating state to avoid further damage and possibly to facilitate recovery once the inflammatory insult has ceased.
